← Back to daily report

api/messages.md

Changed on 2026-09-16 11:17:14 EST

+141 lines added
-0 lines removed
Visual Diff
---¶
title: Messages¶
url: https://platform.claude.com/docs/en/api/messages¶
---¶

# Messages¶

## Create a Message¶

**POST** `/v1/messages`¶

Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.¶

The Messages API can be used for either single queries or stateless multi-turn conversations.¶

Learn more about the Messages API in our [user guide](https://platform.claude.com/docs/en/get-started)¶

### Headers¶

- `"anthropic-user-profile-id": optional string`¶

The user profile ID to attribute this request to. Use when acting on behalf of a party other than your organization. Requires the `user-profiles` beta header.¶

- `"anthropic-workspace-id": optional string`¶

### Body parameters¶

- `max_tokens: number`¶

The maximum number of tokens to generate before stopping.¶

Note that our models may stop _before_ reaching this maximum. This parameter only specifies the absolute maximum number of tokens to generate.¶

Set to `0` to populate the [prompt cache](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#pre-warming-the-cache) without generating a response.¶

Different models have different maximum values for this parameter. See [models](https://platform.claude.com/docs/en/about-claude/models/overview) for details.¶

minimum: 0¶

- `messages: array of MessageParam`¶

Input messages.¶

Our models are trained to operate on alternating `user` and `assistant` conversational turns. When creating a new `Message`, you specify the prior conversational turns with the `messages` parameter, and the model then generates the next `Message` in the conversation. Consecutive `user` or `assistant` turns in your request will be combined into a single turn.¶

Each input message must be an object with a `role` and `content`. You can specify a single `user`-role message, or you can include multiple `user` and `assistant` messages.¶

If the final message uses the `assistant` role, the response content will continue immediately from the content in that message. This can be used to constrain part of the model's response.¶

Example with a single `user` message:¶

```json¶
[{"role": "user", "content": "Hello, Claude"}]¶
```¶

Example with multiple conversational turns:¶

```json¶
[¶
{"role": "user", "content": "Hello there."},¶
{"role": "assistant", "content": "Hi, I'm Claude. How can I help you?"},¶
{"role": "user", "content": "Can you explain LLMs in plain English?"},¶
]¶
```¶

Example with a partially-filled response from Claude:¶

```json¶
[¶
{"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},¶
{"role": "assistant", "content": "The best answer is ("},¶
]¶
```¶

Each input message `content` may be either a single `string` or an array of content blocks, where each block has a specific `type`. Using a `string` for `content` is shorthand for an array of one content block of type `"text"`. The following input messages are equivalent:¶

```json¶
{"role": "user", "content": "Hello, Claude"}¶
```¶

```json¶
{"role": "user", "content": [{"type": "text", "text": "Hello, Claude"}]}¶
```¶

See [input examples](https://platform.claude.com/docs/en/build-with-claude/working-with-messages).¶

Note that if you want to include a [system prompt](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role), you can use the top-level `system` parameter — there is no `"system"` role for input messages in the Messages API.¶

There is a limit of 100,000 messages in a single request.¶

- `content: string or array of ContentBlockParam`¶

- `string`¶

- `array of ContentBlockParam`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

- `DocumentBlockParam object`¶

- `type: "document"`¶

- `source: Base64PDFSource or PlainTextSource or ContentBlockSource or 2 more`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `ContentBlockSource object`¶

- `type: "content"`¶

- `content: string or array of ContentBlockSourceContent`¶

- `string`¶

- `ContentBlockSourceContent = array of ContentBlockSourceContent`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `URLPDFSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileDocumentSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

- `enabled: optional boolean`¶

- `context: optional string or null`¶

minLength: 1¶

- `title: optional string or null`¶

maxLength: 500, minLength: 1¶

- `SearchResultBlockParam object`¶

- `type: "search_result"`¶

- `content: array of TextBlockParam`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional array of TextCitationParam or null`¶

- `source: string`¶

- `title: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam`¶

- `ThinkingBlockParam object`¶

- `type: "thinking"`¶

- `signature: string`¶

The `signature` value of this thinking block, exactly as returned by the API in a previous response. Used to verify that the block was generated by Claude.¶

Thinking blocks must be passed back unmodified and in their original order; a modified block results in a 400 `invalid_request_error`.¶

- `thinking: string`¶

The `thinking` text of this block as returned by the API.¶

- `RedactedThinkingBlockParam object`¶

- `type: "redacted_thinking"`¶

- `data: string`¶

The `data` value of this redacted thinking block, exactly as returned by the API in a previous response. Opaque and encrypted; pass it back unchanged.¶

- `ToolUseBlockParam object`¶

- `type: "tool_use"`¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `input: map[unknown]`¶

- `name: string`¶

maxLength: 200, minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family this member belongs to.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ToolResultBlockParam object`¶

- `type: "tool_result"`¶

- `tool_use_id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `content: optional string or array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `string`¶

- `array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `SearchResultBlockParam object`¶

- `DocumentBlockParam object`¶

- `ToolReferenceBlockParam object`¶

Tool reference block that can be included in tool_result content.¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BrowserStateBlockParam object`¶

The caller's browser state after a browser toolset member call —¶
the full inventory of open tabs, which tab is active, and any side¶
effects (tabs opened, download state changes) the call produced.¶

At most one per `tool_result`, only on a non-error result answering a¶
browser toolset member `tool_use`. The server renders the¶
model-visible text from it; the model never sees the raw fields.¶

- `type: "browser_state"`¶

- `tabs: array of BrowserStateTabEntry`¶

All tabs open in the browser after this call — the full inventory, not a delta. May be empty. Whenever non-empty, exactly one entry carries `active: true`.¶

maxItems: 100¶

- `tab_id: string`¶

The caller-assigned identifier for this tab, unique within the inventory.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `title: string`¶

The title of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The URL of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `active: optional boolean`¶

Whether this tab is the active tab after this call. Whenever `tabs` is non-empty, exactly one entry is marked `active: true`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `state_changes: optional array of BrowserStateChange or null`¶

Tabs opened and download state changes during this call. "Nothing to report" is expressed by omitting the field, never by an empty list.¶

maxItems: 200, minItems: 1¶

- `BrowserStateChangeTabOpened object`¶

A tab this call's execution opened that remains open at its end —¶
the creation delta of the `tabs` inventory, not an event log.¶

Carries only the `tab_id`; the tab's `title` and `url` live on its¶
`tabs` entry, which must include the same `tab_id`. A tab opened¶
during a failed call gets no deferred `tab_opened`; it simply appears¶
in the next result's `tabs` inventory.¶

- `type: "tab_opened"`¶

- `tab_id: string`¶

The `tab_id` of the opened tab, present in `tabs`.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadStarted object`¶

A file download that started during this call.¶

- `type: "download_started"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadCompleted object`¶

A file download that finished during this call, reported with the¶
same `download_id` as its `download_started` — or without a prior¶
`download_started`, when the download finished during the call that¶
started it (at most one state change per `download_id` per result).¶

- `type: "download_completed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `path: optional string or null`¶

Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `size_bytes: optional number or null`¶

The completed download's size.¶

minimum: 0¶

- `BrowserStateChangeDownloadFailed object`¶

A file download that failed — or was cancelled — during this call.¶

- `type: "download_failed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `error: optional string or null`¶

The failure or cancellation detail, when known.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `is_error: optional boolean`¶

- `toolset_name: optional string or null`¶

For a toolset member tool_result, the toolset family of the paired tool_use.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlockParam object`¶

- `type: "server_tool_use"`¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `WebSearchToolResultBlockParam object`¶

- `type: "web_search_tool_result"`¶

- `content: WebSearchToolResultBlockParamContent`¶

- `WebSearchToolResultBlockItem = array of WebSearchResultBlockParam`¶

- `type: "web_search_result"`¶

- `encrypted_content: string`¶

- `title: string`¶

- `url: string`¶

- `page_age: optional string or null`¶

- `WebSearchToolRequestError object`¶

- `type: "web_search_tool_result_error"`¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `WebFetchToolResultBlockParam object`¶

- `type: "web_fetch_tool_result"`¶

- `content: WebFetchToolResultErrorBlockParam or WebFetchBlockParam`¶

- `WebFetchToolResultErrorBlockParam object`¶

- `type: "web_fetch_tool_result_error"`¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlockParam object`¶

- `type: "web_fetch_result"`¶

- `content: DocumentBlockParam`¶

- `url: string`¶

Fetched content URL¶

- `retrieved_at: optional string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `CodeExecutionToolResultBlockParam object`¶

- `type: "code_execution_tool_result"`¶

- `content: CodeExecutionToolResultBlockParamContent`¶

- `CodeExecutionToolResultErrorParam object`¶

- `type: "code_execution_tool_result_error"`¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlockParam object`¶

- `type: "code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlockParam object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BashCodeExecutionToolResultBlockParam object`¶

- `type: "bash_code_execution_tool_result"`¶

- `content: BashCodeExecutionToolResultErrorParam or BashCodeExecutionResultBlockParam`¶

- `BashCodeExecutionToolResultErrorParam object`¶

- `type: "bash_code_execution_tool_result_error"`¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlockParam object`¶

- `type: "bash_code_execution_result"`¶

- `content: array of BashCodeExecutionOutputBlockParam`¶

- `type: "bash_code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `TextEditorCodeExecutionToolResultBlockParam object`¶

- `type: "text_editor_code_execution_tool_result"`¶

- `content: TextEditorCodeExecutionToolResultErrorParam or TextEditorCodeExecutionViewResultBlockParam or TextEditorCodeExecutionCreateResultBlockParam or TextEditorCodeExecutionStrReplaceResultBlockParam`¶

- `TextEditorCodeExecutionToolResultErrorParam object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: optional string or null`¶

- `TextEditorCodeExecutionViewResultBlockParam object`¶

- `type: "text_editor_code_execution_view_result"`¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: optional number or null`¶

- `start_line: optional number or null`¶

- `total_lines: optional number or null`¶

- `TextEditorCodeExecutionCreateResultBlockParam object`¶

- `type: "text_editor_code_execution_create_result"`¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlockParam object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

- `lines: optional array of string or null`¶

- `new_lines: optional number or null`¶

- `new_start: optional number or null`¶

- `old_lines: optional number or null`¶

- `old_start: optional number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `ToolSearchToolResultBlockParam object`¶

- `type: "tool_search_tool_result"`¶

- `content: ToolSearchToolResultErrorParam or ToolSearchToolSearchResultBlockParam`¶

- `ToolSearchToolResultErrorParam object`¶

- `type: "tool_search_tool_result_error"`¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: optional string or null`¶

- `ToolSearchToolSearchResultBlockParam object`¶

- `type: "tool_search_tool_search_result"`¶

- `tool_references: array of ToolReferenceBlockParam`¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `ContainerUploadBlockParam object`¶

A content block that represents a file to be uploaded to the container¶
Files uploaded via this block will be available in the container's input directory.¶

- `type: "container_upload"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `role: "user" or "assistant" or "system"`¶

- `"user"`¶

- `"assistant"`¶

- `"system"`¶

- `model: Model`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1"`¶

Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows¶

- `"claude-mythos-5-1"`¶

Our most capable model for cybersecurity and biology research, available through trusted access programs¶

- `"claude-sonnet-5"`¶

High-performance model for coding and agents¶

- `"claude-fable-5"`¶

Next generation of intelligence for the hardest knowledge work and coding problems¶

- `"claude-mythos-5"`¶

Most capable model for cybersecurity and biology research¶

- `"claude-opus-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-8"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-7"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-mythos-preview"`¶

New class of intelligence, strongest in coding and cybersecurity¶

- `"claude-opus-4-6"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-6"`¶

Best combination of speed and intelligence¶

- `"claude-haiku-4-5"`¶

Fastest model with near-frontier intelligence¶

- `"claude-haiku-4-5-20251001"`¶

Fastest model with near-frontier intelligence¶

- `"claude-opus-4-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-5-20251101"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-5"`¶

High-performance model for agents and coding¶

- `"claude-sonnet-4-5-20250929"`¶

High-performance model for agents and coding¶

- `string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Top-level cache control automatically applies a cache_control marker to the last cacheable block in the request.¶

- `container: optional MessageCreateParamsContainer or null`¶

Container identifier for reuse across requests.¶

- `ContainerParams object`¶

Container parameters with skills to be loaded.¶

- `id: optional string or null`¶

Container id¶

- `skills: optional array of SkillParams or null`¶

List of skills to load in the container¶

maxItems: 20¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: optional string`¶

Skill version or 'latest' for most recent version¶

maxLength: 64, minLength: 1¶

- `string`¶

- `inference_geo: optional string or null`¶

Specifies the geographic region for inference processing. If not specified, the workspace's `default_inference_geo` is used.¶

- `metadata: optional Metadata`¶

An object describing metadata about the request.¶

- `user_id: optional string or null`¶

An external identifier for the user who is associated with the request.¶

This should be a uuid, hash value, or other opaque identifier. Anthropic may use this id to help detect abuse. Do not include any identifying information such as name, email address, or phone number.¶

maxLength: 512¶

- `output_config: optional OutputConfig`¶

Configuration options for the model's output, such as the output format.¶

- `effort: optional "low" or "medium" or "high" or 2 more or null`¶

All possible effort levels.¶

- `"low"`¶

- `"medium"`¶

- `"high"`¶

- `"xhigh"`¶

- `"max"`¶

- `format: optional JSONOutputFormat or null`¶

A schema to specify Claude's output format in responses. See [structured outputs](https://platform.claude.com/docs/en/build-with-claude/structured-outputs)¶

- `type: "json_schema"`¶

- `schema: map[unknown]`¶

The JSON schema of the format¶

- `service_tier: optional "auto" or "standard_only"`¶

Determines whether to use priority capacity (if available) or standard capacity for this request.¶

Anthropic offers different levels of service for your API requests. See [service-tiers](https://platform.claude.com/docs/en/api/service-tiers) for details.¶

- `"auto"`¶

- `"standard_only"`¶

- `stop_sequences: optional array of string`¶

Custom text sequences that will cause the model to stop generating.¶

Our models will normally stop when they have naturally completed their turn, which will result in a response `stop_reason` of `"end_turn"`.¶

If you want the model to stop generating when it encounters custom strings of text, you can use the `stop_sequences` parameter. If the model encounters one of the custom sequences, the response `stop_reason` value will be `"stop_sequence"` and the response `stop_sequence` value will contain the matched stop sequence.¶

- `stream: optional boolean`¶

Whether to incrementally stream the response using server-sent events.¶

See [streaming](https://platform.claude.com/docs/en/build-with-claude/streaming) for details.¶

- `system: optional string or array of TextBlockParam`¶

System prompt.¶

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role. See our [guide to system prompts](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role).¶

- `string`¶

- `array of TextBlockParam`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional array of TextCitationParam or null`¶

- `thinking: optional ThinkingConfigParam`¶

Configuration for enabling Claude's extended thinking.¶

When enabled, responses include `thinking` content blocks showing Claude's thinking process before the final answer. Requires a minimum budget of 1,024 tokens and counts towards your `max_tokens` limit.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `ThinkingConfigEnabled object`¶

- `type: "enabled"`¶

- `budget_tokens: number`¶

Determines how many tokens Claude can use for its internal reasoning process. Larger budgets can enable more thorough analysis for complex problems, improving response quality.¶

Must be ≥1024 and less than `max_tokens`.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

minimum: 1024¶

- `display: optional "summarized" or "omitted" or null`¶

Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.¶

- `"summarized"`¶

- `"omitted"`¶

- `ThinkingConfigDisabled object`¶

- `type: "disabled"`¶

- `ThinkingConfigAdaptive object`¶

- `type: "adaptive"`¶

- `display: optional "summarized" or "omitted" or null`¶

Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.¶

- `"summarized"`¶

- `"omitted"`¶

- `tool_choice: optional ToolChoice`¶

How the model should use the provided tools. The model can use a specific tool, any available tool, decide by itself, or not use tools at all.¶

- `ToolChoiceAuto object`¶

The model will automatically decide whether to use tools.¶

- `type: "auto"`¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output at most one tool use.¶

- `ToolChoiceAny object`¶

The model will use any available tools.¶

- `type: "any"`¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output exactly one tool use.¶

- `ToolChoiceTool object`¶

The model will use the specified tool with `tool_choice.name`.¶

- `type: "tool"`¶

- `name: string`¶

The name of the tool to use.¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output exactly one tool use.¶

- `ToolChoiceNone object`¶

The model will not be allowed to use tools.¶

- `type: "none"`¶

- `tools: optional array of ToolUnion`¶

Definitions of tools that the model may use.¶

If you include `tools` in your API request, the model may return `tool_use` content blocks that represent the model's use of those tools. You can then run those tools using the tool input generated by the model and then optionally return results back to the model using `tool_result` content blocks.¶

There are two types of tools: **client tools** and **server tools**. The behavior described below applies to client tools. For [server tools](https://platform.claude.com/docs/en/agents-and-tools/tool-use/server-tools), see their individual documentation as each has its own behavior (e.g., the [web search tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-search-tool)).¶

Each tool definition includes:¶

* `name`: Name of the tool.¶
* `description`: Optional, but strongly-recommended description of the tool.¶
* `input_schema`: [JSON schema](https://json-schema.org/draft/2020-12) for the tool `input` shape that the model will produce in `tool_use` output content blocks.¶

For example, if you defined `tools` as:¶

```json¶
[¶
{¶
"name": "get_stock_price",¶
"description": "Get the current stock price for a given ticker symbol.",¶
"input_schema": {¶
"type": "object",¶
"properties": {¶
"ticker": {¶
"type": "string",¶
"description": "The stock ticker symbol, e.g. AAPL for Apple Inc."¶
}¶
},¶
"required": ["ticker"]¶
}¶
}¶
]¶
```¶

And then asked the model "What's the S&P 500 at today?", the model might produce `tool_use` content blocks in the response like this:¶

```json¶
[¶
{¶
"type": "tool_use",¶
"id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",¶
"name": "get_stock_price",¶
"input": { "ticker": "^GSPC" }¶
}¶
]¶
```¶

You might then run your `get_stock_price` tool with `{"ticker": "^GSPC"}` as an input, and return the following back to the model in a subsequent `user` message:¶

```json¶
[¶
{¶
"type": "tool_result",¶
"tool_use_id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",¶
"content": "259.75 USD"¶
}¶
]¶
```¶

Tools can be used for workflows that include running client-side tools and functions, or more generally whenever you want the model to produce a particular JSON structure of output.¶

See our [guide](https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview) for more details.¶

- `Tool object`¶

- `type: optional "custom" or null`¶

- `input_schema: object`¶

[JSON schema](https://json-schema.org/draft/2020-12) for this tool's input.¶

This defines the shape of the `input` that your tool accepts and that the model will produce.¶

- `type: "object"`¶

- `properties: optional map[unknown] or null`¶

- `required: optional array of string or null`¶

- `name: string`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

maxLength: 128, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,128}$¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `description: optional string`¶

Description of what this tool does.¶

Tool descriptions should be as detailed as possible. The more information that the model has about what the tool is and how to use it, the better it will perform. You can use natural language descriptions to reinforce important aspects of the tool input JSON schema.¶

- `eager_input_streaming: optional boolean or null`¶

Enable eager input streaming for this tool. When true, tool input parameters will be streamed incrementally as they are generated, and types will be inferred on-the-fly rather than buffering the full JSON output. When false, streaming is disabled for this tool even if the fine-grained-tool-streaming beta is active. When null (default), uses the default behavior based on beta headers.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolBash20250124 object`¶

- `type: "bash_20250124"`¶

- `name: "bash"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20250522 object`¶

- `type: "code_execution_20250522"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20250825 object`¶

- `type: "code_execution_20250825"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20260120 object`¶

Code execution tool with REPL state persistence (daemon mode + gVisor checkpoint).¶

- `type: "code_execution_20260120"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20260521 object`¶

Code execution tool with REPL state persistence.¶

- `type: "code_execution_20260521"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `BrowserToolset20260801 object`¶

The browser toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the browser tool family. The model is served¶
the family's tool with any members disabled via `configs` removed¶
from its schema.¶

- `type: "browser_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `configs: optional BrowserToolsetConfigs or null`¶

Per-member configuration for `browser_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional BrowserTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `close_tab: optional BrowserCloseTabConfig or null`¶

`close_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional BrowserDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `file_upload: optional BrowserFileUploadConfig or null`¶

`file_upload`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `find: optional BrowserFindConfig or null`¶

`find`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `form_input: optional BrowserFormInputConfig or null`¶

`form_input`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `get_page_text: optional BrowserGetPageTextConfig or null`¶

`get_page_text`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional BrowserHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hover: optional BrowserHoverConfig or null`¶

`hover`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `javascript_exec: optional BrowserJavascriptExecConfig or null`¶

`javascript_exec`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional BrowserKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional BrowserLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional BrowserLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional BrowserLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional BrowserLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `list_tabs: optional BrowserListTabsConfig or null`¶

`list_tabs`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional BrowserMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional BrowserMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `navigate: optional BrowserNavigateConfig or null`¶

`navigate`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `new_tab: optional BrowserNewTabConfig or null`¶

`new_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_console: optional BrowserReadConsoleConfig or null`¶

`read_console`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_network: optional BrowserReadNetworkConfig or null`¶

`read_network`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_page: optional BrowserReadPageConfig or null`¶

`read_page`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional BrowserRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional BrowserScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional BrowserScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll_to: optional BrowserScrollToConfig or null`¶

`scroll_to`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `switch_tab: optional BrowserSwitchTabConfig or null`¶

`switch_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional BrowserTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional BrowserWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional BrowserZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `MemoryTool20250818 object`¶

- `type: "memory_20250818"`¶

- `name: "memory"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ComputerToolset20260801 object`¶

The computer toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the computer tool family. The model is¶
served the family's tool with any members disabled via `configs`¶
removed from its schema. Every member is enabled by default, zoom¶
included. The single-tool options `display_number` and¶
`enable_zoom` are not fields of a toolset entry — it carries only¶
`type`, `configs`, and `cache_control`; zoom is controlled¶
via `configs.zoom.enabled`.¶

- `type: "computer_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `configs: optional ComputerToolsetConfigs or null`¶

Per-member configuration for `computer_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional ComputerTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `cursor_position: optional ComputerCursorPositionConfig or null`¶

`cursor_position`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional ComputerDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional ComputerHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional ComputerKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional ComputerLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional ComputerLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional ComputerLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional ComputerLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional ComputerMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional ComputerMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional ComputerRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional ComputerScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional ComputerScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional ComputerTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional ComputerWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional ComputerZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `ToolTextEditor20250124 object`¶

- `type: "text_editor_20250124"`¶

- `name: "str_replace_editor"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolTextEditor20250429 object`¶

- `type: "text_editor_20250429"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolTextEditor20250728 object`¶

- `type: "text_editor_20250728"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `max_characters: optional number or null`¶

Maximum number of characters to display when viewing a file. If not specified, defaults to displaying the full file.¶

minimum: 1¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `WebSearchTool20250305 object`¶

- `type: "web_search_20250305"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `type: "approximate"`¶

- `city: optional string or null`¶

The city of the user.¶

maxLength: 255, minLength: 1¶

- `country: optional string or null`¶

The two letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2) of the user.¶

maxLength: 2, minLength: 2¶

- `region: optional string or null`¶

The region of the user.¶

maxLength: 255, minLength: 1¶

- `timezone: optional string or null`¶

The [IANA timezone](https://nodatime.org/TimeZones) of the user.¶

maxLength: 255, minLength: 1¶

- `WebFetchTool20250910 object`¶

- `type: "web_fetch_20250910"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

- `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`¶

Whether URLs in user messages are fetchable: "all" or "none".¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

 - `WebSearchTool20260209 object`¶

- `type: "web_search_20260209"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `WebFetchTool20260209 object`¶

- `type: "web_fetch_20260209"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `WebFetchTool20260309 object`¶

Web fetch tool with use_cache parameter for bypassing cached content.¶

- `type: "web_fetch_20260309"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

- `WebSearchTool20260318 object`¶

- `type: "web_search_20260318"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `WebFetchTool20260318 object`¶

- `type: "web_fetch_20260318"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

- `ToolSearchToolBm25_20251119 object`¶

- `type: "tool_search_tool_bm25_20251119" or "tool_search_tool_bm25"`¶

- `"tool_search_tool_bm25_20251119"`¶

- `"tool_search_tool_bm25"`¶

- `name: "tool_search_tool_bm25"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolSearchToolRegex20251119 object`¶

- `type: "tool_search_tool_regex_20251119" or "tool_search_tool_regex"`¶

- `"tool_search_tool_regex_20251119"`¶

- `"tool_search_tool_regex"`¶

- `name: "tool_search_tool_regex"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `temperature: optional number`¶

**Deprecated**: Deprecated. Models released after Claude Opus 4.6 do not support setting temperature. A value of 1.0 will be accepted for backwards compatibility, all other values will be rejected with a 400 error.¶

Amount of randomness injected into the response.¶

Defaults to `1.0`. Ranges from `0.0` to `1.0`. Use `temperature` closer to `0.0` for analytical / multiple choice, and closer to `1.0` for creative and generative tasks.¶

Note that even with `temperature` of `0.0`, the results will not be fully deterministic.¶

maximum: 1, minimum: 0¶

- `top_k: optional number`¶

**Deprecated**: Deprecated. Models released after Claude Opus 4.6 do not accept top_k; any value will be rejected with a 400 error.¶

Only sample from the top K options for each subsequent token.¶

Used to remove "long tail" low probability responses. [Learn more technical details here](https://towardsdatascience.com/how-to-sample-from-language-models-682bceb97277).¶

Recommended for advanced use cases only.¶

minimum: 0¶

- `top_p: optional number`¶

**Deprecated**: Deprecated. Models released after Claude Opus 4.6 do not support setting top_p. A value >= 0.99 will be accepted for backwards compatibility, all other values will be rejected with a 400 error.¶

Use nucleus sampling.¶

In nucleus sampling, we compute the cumulative distribution over all the options for each subsequent token in decreasing probability order and cut it off once it reaches a particular probability specified by `top_p`.¶

Recommended for advanced use cases only.¶

maximum: 1, minimum: 0¶

### Returns¶

- `Message object`¶

- `type: "message"`¶

Object type.¶

For Messages, this is always `"message"`.¶

default: message¶

- `id: string`¶

Unique object identifier.¶

The format and length of IDs may change over time.¶

- `container: Container or null`¶

Information about the container used in the request (for the code execution tool)¶

- `id: string`¶

Identifier for the container used in this request¶

- `expires_at: string`¶

The time at which the container will expire.¶

format: date-time¶

- `skills: array of ContainerSkill or null`¶

Skills loaded in the container¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: string`¶

The resolved version: a skill version ID for custom skills.¶

maxLength: 64, minLength: 1¶

- `content: array of ContentBlock`¶

Content generated by the model.¶

This is an array of content blocks, each of which has a `type` that determines its shape.¶

Example:¶

```json¶
[{"type": "text", "text": "Hi, I'm Claude."}]¶
```¶

If the request input `messages` ended with an `assistant` turn, then the response `content` will continue directly from that last turn. You can use this to constrain the model's output.¶

For example, if the input `messages` were:¶

```json¶
[¶
{"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},¶
{"role": "assistant", "content": "The best answer is ("}¶
]¶
```¶

Then the response `content` might be:¶

```json¶
[{"type": "text", "text": "B)"}]¶
```¶

- `TextBlock object`¶

- `type: "text"`¶

default: text¶

- `citations: array of TextCitation or null`¶

Citations supporting the text block.¶

The type of citation returned will depend on the type of document being cited. Citing a PDF results in `page_location`, plain text results in `char_location`, and content document results in `content_block_location`.¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `text: string`¶

minLength: 0¶

- `ThinkingBlock object`¶

- `type: "thinking"`¶

default: thinking¶

- `signature: string`¶

A value used to verify that this thinking block was generated by Claude when it is passed back to the API.¶

This is an opaque field and should not be interpreted or parsed. When passing thinking blocks back to the API (required when using tools with extended thinking), pass them back exactly as received, with this field intact.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `thinking: string`¶

The text of Claude's thinking process for this block.¶

- `RedactedThinkingBlock object`¶

- `type: "redacted_thinking"`¶

default: redacted_thinking¶

- `data: string`¶

The contents of this redacted thinking block, returned when portions of the model's thinking were safety-redacted. This field is opaque and encrypted, with no readable content.¶

Pass `redacted_thinking` blocks back to the API unchanged when continuing a multi-turn conversation.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking#redacted-thinking-blocks) for details.¶

- `ToolUseBlock object`¶

- `type: "tool_use"`¶

default: tool_use¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: string`¶

minLength: 1¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlock object`¶

- `type: "server_tool_use"`¶

default: server_tool_use¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `WebSearchToolResultBlock object`¶

- `type: "web_search_tool_result"`¶

default: web_search_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebSearchToolResultBlockContent`¶

- `WebSearchToolResultError object`¶

- `type: "web_search_tool_result_error"`¶

default: web_search_tool_result_error¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `array of WebSearchResultBlock`¶

- `type: "web_search_result"`¶

default: web_search_result¶

- `encrypted_content: string`¶

- `page_age: string or null`¶

- `title: string`¶

- `url: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `WebFetchToolResultBlock object`¶

- `type: "web_fetch_tool_result"`¶

default: web_fetch_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebFetchToolResultErrorBlock or WebFetchBlock`¶

- `WebFetchToolResultErrorBlock object`¶

- `type: "web_fetch_tool_result_error"`¶

default: web_fetch_tool_result_error¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlock object`¶

- `type: "web_fetch_result"`¶

default: web_fetch_result¶

- `content: DocumentBlock`¶

- `type: "document"`¶

default: document¶

- `citations: CitationsConfig or null`¶

Citation configuration for the document¶

- `enabled: boolean`¶

default: false¶

- `source: Base64PDFSource or PlainTextSource`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `title: string or null`¶

The title of the document¶

- `retrieved_at: string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `url: string`¶

Fetched content URL¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `CodeExecutionToolResultBlock object`¶

- `type: "code_execution_tool_result"`¶

default: code_execution_tool_result¶

- `content: CodeExecutionToolResultBlockContent`¶

- `CodeExecutionToolResultError object`¶

- `type: "code_execution_tool_result_error"`¶

default: code_execution_tool_result_error¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlock object`¶

- `type: "code_execution_result"`¶

default: code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlock object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

default: encrypted_code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `BashCodeExecutionToolResultBlock object`¶

- `type: "bash_code_execution_tool_result"`¶

default: bash_code_execution_tool_result¶

- `content: BashCodeExecutionToolResultError or BashCodeExecutionResultBlock`¶

- `BashCodeExecutionToolResultError object`¶

- `type: "bash_code_execution_tool_result_error"`¶

default: bash_code_execution_tool_result_error¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlock object`¶

- `type: "bash_code_execution_result"`¶

default: bash_code_execution_result¶

- `content: array of BashCodeExecutionOutputBlock`¶

- `type: "bash_code_execution_output"`¶

default: bash_code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `TextEditorCodeExecutionToolResultBlock object`¶

- `type: "text_editor_code_execution_tool_result"`¶

default: text_editor_code_execution_tool_result¶

- `content: TextEditorCodeExecutionToolResultError or TextEditorCodeExecutionViewResultBlock or TextEditorCodeExecutionCreateResultBlock or TextEditorCodeExecutionStrReplaceResultBlock`¶

- `TextEditorCodeExecutionToolResultError object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

default: text_editor_code_execution_tool_result_error¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: string or null`¶

- `TextEditorCodeExecutionViewResultBlock object`¶

- `type: "text_editor_code_execution_view_result"`¶

default: text_editor_code_execution_view_result¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: number or null`¶

- `start_line: number or null`¶

- `total_lines: number or null`¶

- `TextEditorCodeExecutionCreateResultBlock object`¶

- `type: "text_editor_code_execution_create_result"`¶

default: text_editor_code_execution_create_result¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlock object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

default: text_editor_code_execution_str_replace_result¶

- `lines: array of string or null`¶

- `new_lines: number or null`¶

- `new_start: number or null`¶

- `old_lines: number or null`¶

- `old_start: number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ToolSearchToolResultBlock object`¶

- `type: "tool_search_tool_result"`¶

default: tool_search_tool_result¶

- `content: ToolSearchToolResultError or ToolSearchToolSearchResultBlock`¶

- `ToolSearchToolResultError object`¶

- `type: "tool_search_tool_result_error"`¶

default: tool_search_tool_result_error¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: string or null`¶

- `ToolSearchToolSearchResultBlock object`¶

- `type: "tool_search_tool_search_result"`¶

default: tool_search_tool_search_result¶

- `tool_references: array of ToolReferenceBlock`¶

- `type: "tool_reference"`¶

default: tool_reference¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ContainerUploadBlock object`¶

Response model for a file uploaded to the container.¶

- `type: "container_upload"`¶

default: container_upload¶

- `file_id: string`¶

- `model: Model`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1"`¶

Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows¶

- `"claude-mythos-5-1"`¶

Our most capable model for cybersecurity and biology research, available through trusted access programs¶

- `"claude-sonnet-5"`¶

High-performance model for coding and agents¶

- `"claude-fable-5"`¶

Next generation of intelligence for the hardest knowledge work and coding problems¶

- `"claude-mythos-5"`¶

Most capable model for cybersecurity and biology research¶

- `"claude-opus-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-8"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-7"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-mythos-preview"`¶

New class of intelligence, strongest in coding and cybersecurity¶

- `"claude-opus-4-6"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-6"`¶

Best combination of speed and intelligence¶

- `"claude-haiku-4-5"`¶

Fastest model with near-frontier intelligence¶

- `"claude-haiku-4-5-20251001"`¶

Fastest model with near-frontier intelligence¶

- `"claude-opus-4-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-5-20251101"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-5"`¶

High-performance model for agents and coding¶

- `"claude-sonnet-4-5-20250929"`¶

High-performance model for agents and coding¶

- `string`¶

- `role: "assistant"`¶

Conversational role of the generated message.¶

This will always be `"assistant"`.¶

default: assistant¶

- `stop_details: RefusalStopDetails or null`¶

Structured information about a refusal.¶

- `type: "refusal"`¶

default: refusal¶

- `category: "cyber" or "bio" or "frontier_llm" or 2 more or null`¶

The policy category that triggered a refusal.¶

- `"cyber"`¶

The request could enable cyber harm, such as malware or exploit development. Benign cybersecurity work can also trigger this category.¶

- `"bio"`¶

The request could enable biological harm, such as dangerous lab methods. Beneficial life sciences work can also trigger this category.¶

- `"frontier_llm"`¶

The request could assist the development of competing AI models, which is restricted under [Anthropic's commercial terms](https://www.anthropic.com/legal/commercial-terms). Benign machine learning work can also trigger this category.¶

- `"reasoning_extraction"`¶

The request asks the model to reproduce its internal reasoning in the response text. To get reasoning in a structured form instead, use [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/adaptive-thinking).¶

- `"general_harms"`¶

The request could be related to an area that was determined as harmful. Benign work might sometimes trigger this category.¶

- `explanation: string or null`¶

Human-readable explanation of the refusal.¶

This text is not guaranteed to be stable. `null` when no explanation is available for the category.¶

- `stop_reason: StopReason or null`¶

The reason that we stopped.¶

This may be one the following values:¶

* `"end_turn"`: the model reached a natural stopping point¶
* `"max_tokens"`: we exceeded the requested `max_tokens` or the model's maximum¶
* `"stop_sequence"`: one of your provided custom `stop_sequences` was generated¶
* `"tool_use"`: the model invoked one or more tools¶
* `"pause_turn"`: we paused a long-running turn. You may provide the response back as-is in a subsequent request to let the model continue.¶
* `"refusal"`: when streaming classifiers intervene to handle potential policy violations¶
* `"model_context_window_exceeded"`: we exceeded the model's context window¶

In non-streaming mode this value is always non-null. In streaming mode, it is null in the `message_start` event and non-null otherwise.¶

- `"end_turn"`¶

- `"max_tokens"`¶

- `"stop_sequence"`¶

- `"tool_use"`¶

- `"pause_turn"`¶

- `"refusal"`¶

- `"model_context_window_exceeded"`¶

- `stop_sequence: string or null`¶

Which custom stop sequence was generated, if any.¶

This value will be a non-null string if one of your custom stop sequences was generated.¶

- `usage: Usage`¶

Billing and rate-limit usage.¶

Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.¶

Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.¶

For example, `output_tokens` will be non-zero, even for an empty string response from Claude.¶

Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.¶

- `cache_creation: CacheCreation or null`¶

Breakdown of cached tokens by TTL¶

- `ephemeral_1h_input_tokens: number`¶

The number of input tokens used to create the 1 hour cache entry.¶

default: 0, minimum: 0¶

- `ephemeral_5m_input_tokens: number`¶

The number of input tokens used to create the 5 minute cache entry.¶

default: 0, minimum: 0¶

- `cache_creation_input_tokens: number or null`¶

The number of input tokens used to create the cache entry.¶

minimum: 0¶

- `cache_read_input_tokens: number or null`¶

The number of input tokens read from the cache.¶

minimum: 0¶

- `inference_geo: string or null`¶

The geographic region where inference was performed for this request.¶

- `input_tokens: number`¶

The number of input tokens which were used.¶

minimum: 0¶

- `output_tokens: number`¶

The number of output tokens which were used.¶

minimum: 0¶

- `output_tokens_details: OutputTokensDetails or null`¶

Breakdown of output tokens by category.¶

`output_tokens` remains the inclusive, authoritative total used for billing.¶
This object provides a read-only decomposition for observability — for example,¶
how many of the billed output tokens were spent on internal reasoning that may¶
have been summarized before being returned to you.¶

- `thinking_tokens: number`¶

Number of output tokens the model generated as internal reasoning, including¶
the thinking-block delimiter tokens.¶

Reflects the raw reasoning the model produced, not the (possibly shorter)¶
summarized thinking text returned in the response body. Computed by¶
re-tokenizing the raw reasoning text, so it may differ from the model's exact¶
generation count by a small number of tokens. Always ≤ `output_tokens`;¶
`output_tokens - thinking_tokens` approximates the non-reasoning output.¶

default: 0, minimum: 0¶

- `server_tool_use: ServerToolUsage or null`¶

The number of server tool requests.¶

- `web_fetch_requests: number`¶

The number of web fetch tool requests.¶

default: 0, minimum: 0¶

- `web_search_requests: number`¶

The number of web search tool requests.¶

default: 0, minimum: 0¶

- `service_tier: "standard" or "priority" or "batch" or null`¶

If the request used the priority, standard, or batch tier.¶

- `"standard"`¶

- `"priority"`¶

- `"batch"`¶

- `RawMessageStreamEvent = RawMessageStartEvent or RawMessageDeltaEvent or RawMessageStopEvent or 3 more`¶

- `RawMessageStartEvent object`¶

- `type: "message_start"`¶

default: message_start¶

- `message: Message`¶

- `RawMessageDeltaEvent object`¶

- `type: "message_delta"`¶

default: message_delta¶

- `delta: object`¶

- `container: Container or null`¶

Information about the container used in the request (for the code execution tool)¶

- `stop_details: RefusalStopDetails or null`¶

Structured information about a refusal.¶

- `stop_reason: StopReason or null`¶

- `stop_sequence: string or null`¶

- `usage: MessageDeltaUsage`¶

Billing and rate-limit usage.¶

Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.¶

Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.¶

For example, `output_tokens` will be non-zero, even for an empty string response from Claude.¶

Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.¶

- `cache_creation_input_tokens: number or null`¶

The cumulative number of input tokens used to create the cache entry.¶

minimum: 0¶

- `cache_read_input_tokens: number or null`¶

The cumulative number of input tokens read from the cache.¶

minimum: 0¶

- `input_tokens: number or null`¶

The cumulative number of input tokens which were used.¶

minimum: 0¶

- `output_tokens: number`¶

The cumulative number of output tokens which were used.¶

- `output_tokens_details: OutputTokensDetails or null`¶

Breakdown of output tokens by category.¶

`output_tokens` remains the inclusive, authoritative total used for billing.¶
This object provides a read-only decomposition for observability — for example,¶
how many of the billed output tokens were spent on internal reasoning that may¶
have been summarized before being returned to you.¶

- `server_tool_use: ServerToolUsage or null`¶

The number of server tool requests.¶

- `RawMessageStopEvent object`¶

- `type: "message_stop"`¶

default: message_stop¶

- `RawContentBlockStartEvent object`¶

- `type: "content_block_start"`¶

default: content_block_start¶

- `content_block: TextBlock or ThinkingBlock or RedactedThinkingBlock or 9 more`¶

- `TextBlock object`¶

- `ThinkingBlock object`¶

- `RedactedThinkingBlock object`¶

- `ToolUseBlock object`¶

- `ServerToolUseBlock object`¶

- `WebSearchToolResultBlock object`¶

- `WebFetchToolResultBlock object`¶

- `CodeExecutionToolResultBlock object`¶

- `BashCodeExecutionToolResultBlock object`¶

- `TextEditorCodeExecutionToolResultBlock object`¶

- `ToolSearchToolResultBlock object`¶

- `ContainerUploadBlock object`¶

Response model for a file uploaded to the container.¶

- `index: number`¶

- `RawContentBlockDeltaEvent object`¶

- `type: "content_block_delta"`¶

default: content_block_delta¶

- `delta: RawContentBlockDelta`¶

- `TextDelta object`¶

- `type: "text_delta"`¶

default: text_delta¶

- `text: string`¶

- `InputJSONDelta object`¶

- `type: "input_json_delta"`¶

default: input_json_delta¶

- `partial_json: string`¶

- `CitationsDelta object`¶

- `type: "citations_delta"`¶

default: citations_delta¶

- `citation: CitationCharLocation or CitationPageLocation or CitationContentBlockLocation or 2 more`¶

- `CitationCharLocation object`¶

- `CitationPageLocation object`¶

- `CitationContentBlockLocation object`¶

- `CitationsWebSearchResultLocation object`¶

- `CitationsSearchResultLocation object`¶

- `ThinkingDelta object`¶

- `type: "thinking_delta"`¶

default: thinking_delta¶

- `thinking: string`¶

The incremental `thinking` text for this content block. Concatenate the `thinking` values of successive `thinking_delta` events to assemble the block's full `thinking` value.¶

- `SignatureDelta object`¶

- `type: "signature_delta"`¶

default: signature_delta¶

- `signature: string`¶

The `signature` for this thinking block: an opaque value used to verify that the block was generated by Claude when it is passed back to the API. Delivered in a `signature_delta` event just before the block's `content_block_stop` event.¶

- `index: number`¶

- `RawContentBlockStopEvent object`¶

- `type: "content_block_stop"`¶

default: content_block_stop¶

- `index: number`¶

### Example¶

```bash¶
curl https://api.anthropic.com/v1/messages \¶
-H 'Content-Type: application/json' \¶
-H 'anthropic-version: 2023-06-01' \¶
-H "X-Api-Key: $ANTHROPIC_API_KEY" \¶
--max-time 600 \¶
-d '{¶
"max_tokens": 1024,¶
"messages": [¶
{¶
"content": "Hello, world",¶
"role": "user"¶
}¶
],¶
"model": "claude-opus-5",¶
"stream": false,¶
"system": [¶
{¶
"text": "Today'\''s date is 2024-06-01.",¶
"type": "text"¶
}¶
],¶
"temperature": 1,¶
"thinking": {¶
"type": "adaptive"¶
},¶
"tools": [¶
{¶
"input_schema": {¶
"type": "object",¶
"properties": {¶
"location": "bar",¶
"unit": "bar"¶
},¶
"required": [¶
"location"¶
]¶
},¶
"name": "name"¶
}¶
],¶
"top_k": 5,¶
"top_p": 0.7¶
}'¶
```¶

#### Response (200)¶

```json¶
{¶
"id": "msg_013Zva2CMHLNnXjNJJKqJ2EF",¶
"container": {¶
"id": "container_011CpZohnwH4vuy7gazohgSP",¶
"expires_at": "2019-12-27T18:11:19.117Z",¶
"skills": [¶
{¶
"skill_id": "pdf",¶
"type": "anthropic",¶
"version": "latest"¶
}¶
]¶
},¶
"content": [¶
{¶
"citations": [¶
{¶
"cited_text": "The grass is green. The sky is blue.",¶
"document_index": 0,¶
"document_title": "My Document",¶
"end_char_index": 0,¶
"file_id": "file_011CNha8iCJcU1wXNR6q4V8w",¶
"start_char_index": 0,¶
"type": "char_location"¶
}¶
],¶
"text": "Hi! My name is Claude.",¶
"type": "text"¶
}¶
],¶
"model": "claude-opus-5",¶
"role": "assistant",¶
"stop_details": {¶
"category": "cyber",¶
"explanation": "This request was declined because it conflicts with Anthropic's Usage Policy.",¶
"type": "refusal"¶
},¶
"stop_reason": "end_turn",¶
"stop_sequence": null,¶
"type": "message",¶
"usage": {¶
"cache_creation": {¶
"ephemeral_1h_input_tokens": 0,¶
"ephemeral_5m_input_tokens": 0¶
},¶
"cache_creation_input_tokens": 2051,¶
"cache_read_input_tokens": 2051,¶
"inference_geo": "global",¶
"input_tokens": 2095,¶
"output_tokens": 503,¶
"output_tokens_details": {¶
"thinking_tokens": 0¶
},¶
"server_tool_use": {¶
"web_fetch_requests": 2,¶
"web_search_requests": 0¶
},¶
"service_tier": "standard"¶
}¶
}¶
```¶

## Count tokens in a Message¶

**POST** `/v1/messages/count_tokens`¶

Count the number of tokens in a Message.¶

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.¶

Learn more about token counting in our [user guide](https://platform.claude.com/docs/en/build-with-claude/token-counting)¶

### Headers¶

- `"anthropic-user-profile-id": optional string`¶

The user profile ID to attribute this request to. Use when acting on behalf of a party other than your organization. Requires the `user-profiles` beta header.¶

- `"anthropic-workspace-id": optional string`¶

### Body parameters¶

- `messages: array of MessageParam`¶

Input messages.¶

Our models are trained to operate on alternating `user` and `assistant` conversational turns. When creating a new `Message`, you specify the prior conversational turns with the `messages` parameter, and the model then generates the next `Message` in the conversation. Consecutive `user` or `assistant` turns in your request will be combined into a single turn.¶

Each input message must be an object with a `role` and `content`. You can specify a single `user`-role message, or you can include multiple `user` and `assistant` messages.¶

If the final message uses the `assistant` role, the response content will continue immediately from the content in that message. This can be used to constrain part of the model's response.¶

Example with a single `user` message:¶

```json¶
[{"role": "user", "content": "Hello, Claude"}]¶
```¶

Example with multiple conversational turns:¶

```json¶
[¶
{"role": "user", "content": "Hello there."},¶
{"role": "assistant", "content": "Hi, I'm Claude. How can I help you?"},¶
{"role": "user", "content": "Can you explain LLMs in plain English?"},¶
]¶
```¶

Example with a partially-filled response from Claude:¶

```json¶
[¶
{"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},¶
{"role": "assistant", "content": "The best answer is ("},¶
]¶
```¶

Each input message `content` may be either a single `string` or an array of content blocks, where each block has a specific `type`. Using a `string` for `content` is shorthand for an array of one content block of type `"text"`. The following input messages are equivalent:¶

```json¶
{"role": "user", "content": "Hello, Claude"}¶
```¶

```json¶
{"role": "user", "content": [{"type": "text", "text": "Hello, Claude"}]}¶
```¶

See [input examples](https://platform.claude.com/docs/en/build-with-claude/working-with-messages).¶

Note that if you want to include a [system prompt](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role), you can use the top-level `system` parameter — there is no `"system"` role for input messages in the Messages API.¶

There is a limit of 100,000 messages in a single request.¶

- `content: string or array of ContentBlockParam`¶

- `string`¶

- `array of ContentBlockParam`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

- `DocumentBlockParam object`¶

- `type: "document"`¶

- `source: Base64PDFSource or PlainTextSource or ContentBlockSource or 2 more`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `ContentBlockSource object`¶

- `type: "content"`¶

- `content: string or array of ContentBlockSourceContent`¶

- `string`¶

- `ContentBlockSourceContent = array of ContentBlockSourceContent`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `URLPDFSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileDocumentSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

- `enabled: optional boolean`¶

- `context: optional string or null`¶

minLength: 1¶

- `title: optional string or null`¶

maxLength: 500, minLength: 1¶

- `SearchResultBlockParam object`¶

- `type: "search_result"`¶

- `content: array of TextBlockParam`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional array of TextCitationParam or null`¶

- `source: string`¶

- `title: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam`¶

- `ThinkingBlockParam object`¶

- `type: "thinking"`¶

- `signature: string`¶

The `signature` value of this thinking block, exactly as returned by the API in a previous response. Used to verify that the block was generated by Claude.¶

Thinking blocks must be passed back unmodified and in their original order; a modified block results in a 400 `invalid_request_error`.¶

- `thinking: string`¶

The `thinking` text of this block as returned by the API.¶

- `RedactedThinkingBlockParam object`¶

- `type: "redacted_thinking"`¶

- `data: string`¶

The `data` value of this redacted thinking block, exactly as returned by the API in a previous response. Opaque and encrypted; pass it back unchanged.¶

- `ToolUseBlockParam object`¶

- `type: "tool_use"`¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `input: map[unknown]`¶

- `name: string`¶

maxLength: 200, minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family this member belongs to.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ToolResultBlockParam object`¶

- `type: "tool_result"`¶

- `tool_use_id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `content: optional string or array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `string`¶

- `array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `SearchResultBlockParam object`¶

- `DocumentBlockParam object`¶

- `ToolReferenceBlockParam object`¶

Tool reference block that can be included in tool_result content.¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BrowserStateBlockParam object`¶

The caller's browser state after a browser toolset member call —¶
the full inventory of open tabs, which tab is active, and any side¶
effects (tabs opened, download state changes) the call produced.¶

At most one per `tool_result`, only on a non-error result answering a¶
browser toolset member `tool_use`. The server renders the¶
model-visible text from it; the model never sees the raw fields.¶

- `type: "browser_state"`¶

- `tabs: array of BrowserStateTabEntry`¶

All tabs open in the browser after this call — the full inventory, not a delta. May be empty. Whenever non-empty, exactly one entry carries `active: true`.¶

maxItems: 100¶

- `tab_id: string`¶

The caller-assigned identifier for this tab, unique within the inventory.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `title: string`¶

The title of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The URL of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `active: optional boolean`¶

Whether this tab is the active tab after this call. Whenever `tabs` is non-empty, exactly one entry is marked `active: true`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `state_changes: optional array of BrowserStateChange or null`¶

Tabs opened and download state changes during this call. "Nothing to report" is expressed by omitting the field, never by an empty list.¶

maxItems: 200, minItems: 1¶

- `BrowserStateChangeTabOpened object`¶

A tab this call's execution opened that remains open at its end —¶
the creation delta of the `tabs` inventory, not an event log.¶

Carries only the `tab_id`; the tab's `title` and `url` live on its¶
`tabs` entry, which must include the same `tab_id`. A tab opened¶
during a failed call gets no deferred `tab_opened`; it simply appears¶
in the next result's `tabs` inventory.¶

- `type: "tab_opened"`¶

- `tab_id: string`¶

The `tab_id` of the opened tab, present in `tabs`.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadStarted object`¶

A file download that started during this call.¶

- `type: "download_started"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadCompleted object`¶

A file download that finished during this call, reported with the¶
same `download_id` as its `download_started` — or without a prior¶
`download_started`, when the download finished during the call that¶
started it (at most one state change per `download_id` per result).¶

- `type: "download_completed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `path: optional string or null`¶

Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `size_bytes: optional number or null`¶

The completed download's size.¶

minimum: 0¶

- `BrowserStateChangeDownloadFailed object`¶

A file download that failed — or was cancelled — during this call.¶

- `type: "download_failed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `error: optional string or null`¶

The failure or cancellation detail, when known.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `is_error: optional boolean`¶

- `toolset_name: optional string or null`¶

For a toolset member tool_result, the toolset family of the paired tool_use.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlockParam object`¶

- `type: "server_tool_use"`¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `WebSearchToolResultBlockParam object`¶

- `type: "web_search_tool_result"`¶

- `content: WebSearchToolResultBlockParamContent`¶

- `WebSearchToolResultBlockItem = array of WebSearchResultBlockParam`¶

- `type: "web_search_result"`¶

- `encrypted_content: string`¶

- `title: string`¶

- `url: string`¶

- `page_age: optional string or null`¶

- `WebSearchToolRequestError object`¶

- `type: "web_search_tool_result_error"`¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `WebFetchToolResultBlockParam object`¶

- `type: "web_fetch_tool_result"`¶

- `content: WebFetchToolResultErrorBlockParam or WebFetchBlockParam`¶

- `WebFetchToolResultErrorBlockParam object`¶

- `type: "web_fetch_tool_result_error"`¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlockParam object`¶

- `type: "web_fetch_result"`¶

- `content: DocumentBlockParam`¶

- `url: string`¶

Fetched content URL¶

- `retrieved_at: optional string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `CodeExecutionToolResultBlockParam object`¶

- `type: "code_execution_tool_result"`¶

- `content: CodeExecutionToolResultBlockParamContent`¶

- `CodeExecutionToolResultErrorParam object`¶

- `type: "code_execution_tool_result_error"`¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlockParam object`¶

- `type: "code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlockParam object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BashCodeExecutionToolResultBlockParam object`¶

- `type: "bash_code_execution_tool_result"`¶

- `content: BashCodeExecutionToolResultErrorParam or BashCodeExecutionResultBlockParam`¶

- `BashCodeExecutionToolResultErrorParam object`¶

- `type: "bash_code_execution_tool_result_error"`¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlockParam object`¶

- `type: "bash_code_execution_result"`¶

- `content: array of BashCodeExecutionOutputBlockParam`¶

- `type: "bash_code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `TextEditorCodeExecutionToolResultBlockParam object`¶

- `type: "text_editor_code_execution_tool_result"`¶

- `content: TextEditorCodeExecutionToolResultErrorParam or TextEditorCodeExecutionViewResultBlockParam or TextEditorCodeExecutionCreateResultBlockParam or TextEditorCodeExecutionStrReplaceResultBlockParam`¶

- `TextEditorCodeExecutionToolResultErrorParam object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: optional string or null`¶

- `TextEditorCodeExecutionViewResultBlockParam object`¶

- `type: "text_editor_code_execution_view_result"`¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: optional number or null`¶

- `start_line: optional number or null`¶

- `total_lines: optional number or null`¶

- `TextEditorCodeExecutionCreateResultBlockParam object`¶

- `type: "text_editor_code_execution_create_result"`¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlockParam object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

- `lines: optional array of string or null`¶

- `new_lines: optional number or null`¶

- `new_start: optional number or null`¶

- `old_lines: optional number or null`¶

- `old_start: optional number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `ToolSearchToolResultBlockParam object`¶

- `type: "tool_search_tool_result"`¶

- `content: ToolSearchToolResultErrorParam or ToolSearchToolSearchResultBlockParam`¶

- `ToolSearchToolResultErrorParam object`¶

- `type: "tool_search_tool_result_error"`¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: optional string or null`¶

- `ToolSearchToolSearchResultBlockParam object`¶

- `type: "tool_search_tool_search_result"`¶

- `tool_references: array of ToolReferenceBlockParam`¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `ContainerUploadBlockParam object`¶

A content block that represents a file to be uploaded to the container¶
Files uploaded via this block will be available in the container's input directory.¶

- `type: "container_upload"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `role: "user" or "assistant" or "system"`¶

- `"user"`¶

- `"assistant"`¶

- `"system"`¶

- `model: Model`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1"`¶

Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows¶

- `"claude-mythos-5-1"`¶

Our most capable model for cybersecurity and biology research, available through trusted access programs¶

- `"claude-sonnet-5"`¶

High-performance model for coding and agents¶

- `"claude-fable-5"`¶

Next generation of intelligence for the hardest knowledge work and coding problems¶

- `"claude-mythos-5"`¶

Most capable model for cybersecurity and biology research¶

- `"claude-opus-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-8"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-7"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-mythos-preview"`¶

New class of intelligence, strongest in coding and cybersecurity¶

- `"claude-opus-4-6"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-6"`¶

Best combination of speed and intelligence¶

- `"claude-haiku-4-5"`¶

Fastest model with near-frontier intelligence¶

- `"claude-haiku-4-5-20251001"`¶

Fastest model with near-frontier intelligence¶

- `"claude-opus-4-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-5-20251101"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-5"`¶

High-performance model for agents and coding¶

- `"claude-sonnet-4-5-20250929"`¶

High-performance model for agents and coding¶

- `string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Top-level cache control automatically applies a cache_control marker to the last cacheable block in the request.¶

- `output_config: optional OutputConfig`¶

Configuration options for the model's output, such as the output format.¶

- `effort: optional "low" or "medium" or "high" or 2 more or null`¶

All possible effort levels.¶

- `"low"`¶

- `"medium"`¶

- `"high"`¶

- `"xhigh"`¶

- `"max"`¶

- `format: optional JSONOutputFormat or null`¶

A schema to specify Claude's output format in responses. See [structured outputs](https://platform.claude.com/docs/en/build-with-claude/structured-outputs)¶

- `type: "json_schema"`¶

- `schema: map[unknown]`¶

The JSON schema of the format¶

- `system: optional string or array of TextBlockParam`¶

System prompt.¶

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role. See our [guide to system prompts](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role).¶

- `string`¶

- `array of TextBlockParam`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional array of TextCitationParam or null`¶

- `thinking: optional ThinkingConfigParam`¶

Configuration for enabling Claude's extended thinking.¶

When enabled, responses include `thinking` content blocks showing Claude's thinking process before the final answer. Requires a minimum budget of 1,024 tokens and counts towards your `max_tokens` limit.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `ThinkingConfigEnabled object`¶

- `type: "enabled"`¶

- `budget_tokens: number`¶

Determines how many tokens Claude can use for its internal reasoning process. Larger budgets can enable more thorough analysis for complex problems, improving response quality.¶

Must be ≥1024 and less than `max_tokens`.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

minimum: 1024¶

- `display: optional "summarized" or "omitted" or null`¶

Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.¶

- `"summarized"`¶

- `"omitted"`¶

- `ThinkingConfigDisabled object`¶

- `type: "disabled"`¶

- `ThinkingConfigAdaptive object`¶

- `type: "adaptive"`¶

- `display: optional "summarized" or "omitted" or null`¶

Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.¶

- `"summarized"`¶

- `"omitted"`¶

- `tool_choice: optional ToolChoice`¶

How the model should use the provided tools. The model can use a specific tool, any available tool, decide by itself, or not use tools at all.¶

- `ToolChoiceAuto object`¶

The model will automatically decide whether to use tools.¶

- `type: "auto"`¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output at most one tool use.¶

- `ToolChoiceAny object`¶

The model will use any available tools.¶

- `type: "any"`¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output exactly one tool use.¶

- `ToolChoiceTool object`¶

The model will use the specified tool with `tool_choice.name`.¶

- `type: "tool"`¶

- `name: string`¶

The name of the tool to use.¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output exactly one tool use.¶

- `ToolChoiceNone object`¶

The model will not be allowed to use tools.¶

- `type: "none"`¶

- `tools: optional array of MessageCountTokensTool`¶

Definitions of tools that the model may use.¶

If you include `tools` in your API request, the model may return `tool_use` content blocks that represent the model's use of those tools. You can then run those tools using the tool input generated by the model and then optionally return results back to the model using `tool_result` content blocks.¶

There are two types of tools: **client tools** and **server tools**. The behavior described below applies to client tools. For [server tools](https://platform.claude.com/docs/en/agents-and-tools/tool-use/server-tools), see their individual documentation as each has its own behavior (e.g., the [web search tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-search-tool)).¶

Each tool definition includes:¶

* `name`: Name of the tool.¶
* `description`: Optional, but strongly-recommended description of the tool.¶
* `input_schema`: [JSON schema](https://json-schema.org/draft/2020-12) for the tool `input` shape that the model will produce in `tool_use` output content blocks.¶

For example, if you defined `tools` as:¶

```json¶
[¶
{¶
"name": "get_stock_price",¶
"description": "Get the current stock price for a given ticker symbol.",¶
"input_schema": {¶
"type": "object",¶
"properties": {¶
"ticker": {¶
"type": "string",¶
"description": "The stock ticker symbol, e.g. AAPL for Apple Inc."¶
}¶
},¶
"required": ["ticker"]¶
}¶
}¶
]¶
```¶

And then asked the model "What's the S&P 500 at today?", the model might produce `tool_use` content blocks in the response like this:¶

```json¶
[¶
{¶
"type": "tool_use",¶
"id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",¶
"name": "get_stock_price",¶
"input": { "ticker": "^GSPC" }¶
}¶
]¶
```¶

You might then run your `get_stock_price` tool with `{"ticker": "^GSPC"}` as an input, and return the following back to the model in a subsequent `user` message:¶

```json¶
[¶
{¶
"type": "tool_result",¶
"tool_use_id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",¶
"content": "259.75 USD"¶
}¶
]¶
```¶

Tools can be used for workflows that include running client-side tools and functions, or more generally whenever you want the model to produce a particular JSON structure of output.¶

See our [guide](https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview) for more details.¶

- `Tool object`¶

- `type: optional "custom" or null`¶

- `input_schema: object`¶

[JSON schema](https://json-schema.org/draft/2020-12) for this tool's input.¶

This defines the shape of the `input` that your tool accepts and that the model will produce.¶

- `type: "object"`¶

- `properties: optional map[unknown] or null`¶

- `required: optional array of string or null`¶

- `name: string`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

maxLength: 128, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,128}$¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `description: optional string`¶

Description of what this tool does.¶

Tool descriptions should be as detailed as possible. The more information that the model has about what the tool is and how to use it, the better it will perform. You can use natural language descriptions to reinforce important aspects of the tool input JSON schema.¶

- `eager_input_streaming: optional boolean or null`¶

Enable eager input streaming for this tool. When true, tool input parameters will be streamed incrementally as they are generated, and types will be inferred on-the-fly rather than buffering the full JSON output. When false, streaming is disabled for this tool even if the fine-grained-tool-streaming beta is active. When null (default), uses the default behavior based on beta headers.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolBash20250124 object`¶

- `type: "bash_20250124"`¶

- `name: "bash"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20250522 object`¶

- `type: "code_execution_20250522"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20250825 object`¶

- `type: "code_execution_20250825"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20260120 object`¶

Code execution tool with REPL state persistence (daemon mode + gVisor checkpoint).¶

- `type: "code_execution_20260120"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20260521 object`¶

Code execution tool with REPL state persistence.¶

- `type: "code_execution_20260521"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `BrowserToolset20260801 object`¶

The browser toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the browser tool family. The model is served¶
the family's tool with any members disabled via `configs` removed¶
from its schema.¶

- `type: "browser_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `configs: optional BrowserToolsetConfigs or null`¶

Per-member configuration for `browser_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional BrowserTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `close_tab: optional BrowserCloseTabConfig or null`¶

`close_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional BrowserDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `file_upload: optional BrowserFileUploadConfig or null`¶

`file_upload`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `find: optional BrowserFindConfig or null`¶

`find`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `form_input: optional BrowserFormInputConfig or null`¶

`form_input`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `get_page_text: optional BrowserGetPageTextConfig or null`¶

`get_page_text`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional BrowserHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hover: optional BrowserHoverConfig or null`¶

`hover`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `javascript_exec: optional BrowserJavascriptExecConfig or null`¶

`javascript_exec`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional BrowserKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional BrowserLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional BrowserLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional BrowserLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional BrowserLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `list_tabs: optional BrowserListTabsConfig or null`¶

`list_tabs`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional BrowserMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional BrowserMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `navigate: optional BrowserNavigateConfig or null`¶

`navigate`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `new_tab: optional BrowserNewTabConfig or null`¶

`new_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_console: optional BrowserReadConsoleConfig or null`¶

`read_console`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_network: optional BrowserReadNetworkConfig or null`¶

`read_network`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_page: optional BrowserReadPageConfig or null`¶

`read_page`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional BrowserRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional BrowserScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional BrowserScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll_to: optional BrowserScrollToConfig or null`¶

`scroll_to`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `switch_tab: optional BrowserSwitchTabConfig or null`¶

`switch_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional BrowserTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional BrowserWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional BrowserZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `MemoryTool20250818 object`¶

- `type: "memory_20250818"`¶

- `name: "memory"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ComputerToolset20260801 object`¶

The computer toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the computer tool family. The model is¶
served the family's tool with any members disabled via `configs`¶
removed from its schema. Every member is enabled by default, zoom¶
included. The single-tool options `display_number` and¶
`enable_zoom` are not fields of a toolset entry — it carries only¶
`type`, `configs`, and `cache_control`; zoom is controlled¶
via `configs.zoom.enabled`.¶

- `type: "computer_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `configs: optional ComputerToolsetConfigs or null`¶

Per-member configuration for `computer_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional ComputerTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `cursor_position: optional ComputerCursorPositionConfig or null`¶

`cursor_position`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional ComputerDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional ComputerHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional ComputerKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional ComputerLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional ComputerLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional ComputerLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional ComputerLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional ComputerMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional ComputerMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional ComputerRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional ComputerScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional ComputerScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional ComputerTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional ComputerWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional ComputerZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `ToolTextEditor20250124 object`¶

- `type: "text_editor_20250124"`¶

- `name: "str_replace_editor"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolTextEditor20250429 object`¶

- `type: "text_editor_20250429"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolTextEditor20250728 object`¶

- `type: "text_editor_20250728"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `max_characters: optional number or null`¶

Maximum number of characters to display when viewing a file. If not specified, defaults to displaying the full file.¶

minimum: 1¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `WebSearchTool20250305 object`¶

- `type: "web_search_20250305"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `type: "approximate"`¶

- `city: optional string or null`¶

The city of the user.¶

maxLength: 255, minLength: 1¶

- `country: optional string or null`¶

The two letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2) of the user.¶

maxLength: 2, minLength: 2¶

- `region: optional string or null`¶

The region of the user.¶

maxLength: 255, minLength: 1¶

- `timezone: optional string or null`¶

The [IANA timezone](https://nodatime.org/TimeZones) of the user.¶

maxLength: 255, minLength: 1¶

- `WebFetchTool20250910 object`¶

- `type: "web_fetch_20250910"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

- `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`¶

Whether URLs in user messages are fetchable: "all" or "none".¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

 - `WebSearchTool20260209 object`¶

- `type: "web_search_20260209"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `WebFetchTool20260209 object`¶

- `type: "web_fetch_20260209"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `WebFetchTool20260309 object`¶

Web fetch tool with use_cache parameter for bypassing cached content.¶

- `type: "web_fetch_20260309"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

- `WebSearchTool20260318 object`¶

- `type: "web_search_20260318"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `WebFetchTool20260318 object`¶

- `type: "web_fetch_20260318"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

- `ToolSearchToolBm25_20251119 object`¶

- `type: "tool_search_tool_bm25_20251119" or "tool_search_tool_bm25"`¶

- `"tool_search_tool_bm25_20251119"`¶

- `"tool_search_tool_bm25"`¶

- `name: "tool_search_tool_bm25"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolSearchToolRegex20251119 object`¶

- `type: "tool_search_tool_regex_20251119" or "tool_search_tool_regex"`¶

- `"tool_search_tool_regex_20251119"`¶

- `"tool_search_tool_regex"`¶

- `name: "tool_search_tool_regex"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Returns¶

- `MessageTokensCount object`¶

- `input_tokens: number`¶

The total number of tokens across the provided list of messages, system prompt, and tools.¶

### Example¶

```bash¶
curl https://api.anthropic.com/v1/messages/count_tokens \¶
-H 'Content-Type: application/json' \¶
-H 'anthropic-version: 2023-06-01' \¶
-H "X-Api-Key: $ANTHROPIC_API_KEY" \¶
-d '{¶
"messages": [¶
{¶
"content": "Hello, world",¶
"role": "user"¶
}¶
],¶
"model": "claude-opus-5",¶
"system": [¶
{¶
"text": "Today'\''s date is 2024-06-01.",¶
"type": "text"¶
}¶
],¶
"thinking": {¶
"type": "adaptive"¶
},¶
"tools": [¶
{¶
"input_schema": {¶
"type": "object",¶
"properties": {¶
"location": "bar",¶
"unit": "bar"¶
},¶
"required": [¶
"location"¶
]¶
},¶
"name": "name"¶
}¶
]¶
}'¶
```¶

#### Response (200)¶

```json¶
{¶
"input_tokens": 2095¶
}¶
```¶

## Domain types¶

### Base64 Image Source¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

### Base64 PDF Source¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

### Bash Code Execution Output Block¶

- `BashCodeExecutionOutputBlock object`¶

- `type: "bash_code_execution_output"`¶

default: bash_code_execution_output¶

- `file_id: string`¶

### Bash Code Execution Output Block Param¶

- `BashCodeExecutionOutputBlockParam object`¶

- `type: "bash_code_execution_output"`¶

- `file_id: string`¶

### Bash Code Execution Result Block¶

- `BashCodeExecutionResultBlock object`¶

- `type: "bash_code_execution_result"`¶

default: bash_code_execution_result¶

- `content: array of BashCodeExecutionOutputBlock`¶

- `type: "bash_code_execution_output"`¶

default: bash_code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

### Bash Code Execution Result Block Param¶

- `BashCodeExecutionResultBlockParam object`¶

- `type: "bash_code_execution_result"`¶

- `content: array of BashCodeExecutionOutputBlockParam`¶

- `type: "bash_code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

### Bash Code Execution Tool Result Block¶

- `BashCodeExecutionToolResultBlock object`¶

- `type: "bash_code_execution_tool_result"`¶

default: bash_code_execution_tool_result¶

- `content: BashCodeExecutionToolResultError or BashCodeExecutionResultBlock`¶

- `BashCodeExecutionToolResultError object`¶

- `type: "bash_code_execution_tool_result_error"`¶

default: bash_code_execution_tool_result_error¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlock object`¶

- `type: "bash_code_execution_result"`¶

default: bash_code_execution_result¶

- `content: array of BashCodeExecutionOutputBlock`¶

- `type: "bash_code_execution_output"`¶

default: bash_code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Bash Code Execution Tool Result Block Param¶

- `BashCodeExecutionToolResultBlockParam object`¶

- `type: "bash_code_execution_tool_result"`¶

- `content: BashCodeExecutionToolResultErrorParam or BashCodeExecutionResultBlockParam`¶

- `BashCodeExecutionToolResultErrorParam object`¶

- `type: "bash_code_execution_tool_result_error"`¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlockParam object`¶

- `type: "bash_code_execution_result"`¶

- `content: array of BashCodeExecutionOutputBlockParam`¶

- `type: "bash_code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

### Bash Code Execution Tool Result Error¶

- `BashCodeExecutionToolResultError object`¶

- `type: "bash_code_execution_tool_result_error"`¶

default: bash_code_execution_tool_result_error¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

### Bash Code Execution Tool Result Error Code¶

- `BashCodeExecutionToolResultErrorCode = "invalid_tool_input" or "unavailable" or "too_many_requests" or 2 more`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

### Bash Code Execution Tool Result Error Param¶

- `BashCodeExecutionToolResultErrorParam object`¶

- `type: "bash_code_execution_tool_result_error"`¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

### Browser Close Tab Config¶

- `BrowserCloseTabConfig object`¶

`close_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Double Click Config¶

- `BrowserDoubleClickConfig object`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser File Upload Config¶

- `BrowserFileUploadConfig object`¶

`file_upload`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Find Config¶

- `BrowserFindConfig object`¶

`find`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Form Input Config¶

- `BrowserFormInputConfig object`¶

`form_input`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Get Page Text Config¶

- `BrowserGetPageTextConfig object`¶

`get_page_text`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Hold Key Config¶

- `BrowserHoldKeyConfig object`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Hover Config¶

- `BrowserHoverConfig object`¶

`hover`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Javascript Exec Config¶

- `BrowserJavascriptExecConfig object`¶

`javascript_exec`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Key Config¶

- `BrowserKeyConfig object`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Left Click Config¶

- `BrowserLeftClickConfig object`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Left Click Drag Config¶

- `BrowserLeftClickDragConfig object`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Left Mouse Down Config¶

- `BrowserLeftMouseDownConfig object`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Left Mouse Up Config¶

- `BrowserLeftMouseUpConfig object`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser List Tabs Config¶

- `BrowserListTabsConfig object`¶

`list_tabs`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Middle Click Config¶

- `BrowserMiddleClickConfig object`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Mouse Move Config¶

- `BrowserMouseMoveConfig object`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Navigate Config¶

- `BrowserNavigateConfig object`¶

`navigate`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser New Tab Config¶

- `BrowserNewTabConfig object`¶

`new_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Read Console Config¶

- `BrowserReadConsoleConfig object`¶

`read_console`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Read Network Config¶

- `BrowserReadNetworkConfig object`¶

`read_network`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Read Page Config¶

- `BrowserReadPageConfig object`¶

`read_page`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Right Click Config¶

- `BrowserRightClickConfig object`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Screenshot Config¶

- `BrowserScreenshotConfig object`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Scroll Config¶

- `BrowserScrollConfig object`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Scroll To Config¶

- `BrowserScrollToConfig object`¶

`scroll_to`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser State Block Param¶

- `BrowserStateBlockParam object`¶

The caller's browser state after a browser toolset member call —¶
the full inventory of open tabs, which tab is active, and any side¶
effects (tabs opened, download state changes) the call produced.¶

At most one per `tool_result`, only on a non-error result answering a¶
browser toolset member `tool_use`. The server renders the¶
model-visible text from it; the model never sees the raw fields.¶

- `type: "browser_state"`¶

- `tabs: array of BrowserStateTabEntry`¶

All tabs open in the browser after this call — the full inventory, not a delta. May be empty. Whenever non-empty, exactly one entry carries `active: true`.¶

maxItems: 100¶

- `tab_id: string`¶

The caller-assigned identifier for this tab, unique within the inventory.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `title: string`¶

The title of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The URL of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `active: optional boolean`¶

Whether this tab is the active tab after this call. Whenever `tabs` is non-empty, exactly one entry is marked `active: true`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `state_changes: optional array of BrowserStateChange or null`¶

Tabs opened and download state changes during this call. "Nothing to report" is expressed by omitting the field, never by an empty list.¶

maxItems: 200, minItems: 1¶

- `BrowserStateChangeTabOpened object`¶

A tab this call's execution opened that remains open at its end —¶
the creation delta of the `tabs` inventory, not an event log.¶

Carries only the `tab_id`; the tab's `title` and `url` live on its¶
`tabs` entry, which must include the same `tab_id`. A tab opened¶
during a failed call gets no deferred `tab_opened`; it simply appears¶
in the next result's `tabs` inventory.¶

- `type: "tab_opened"`¶

- `tab_id: string`¶

The `tab_id` of the opened tab, present in `tabs`.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadStarted object`¶

A file download that started during this call.¶

- `type: "download_started"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadCompleted object`¶

A file download that finished during this call, reported with the¶
same `download_id` as its `download_started` — or without a prior¶
`download_started`, when the download finished during the call that¶
started it (at most one state change per `download_id` per result).¶

- `type: "download_completed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `path: optional string or null`¶

Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `size_bytes: optional number or null`¶

The completed download's size.¶

minimum: 0¶

- `BrowserStateChangeDownloadFailed object`¶

A file download that failed — or was cancelled — during this call.¶

- `type: "download_failed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `error: optional string or null`¶

The failure or cancellation detail, when known.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

### Browser State Change¶

- `BrowserStateChange = BrowserStateChangeTabOpened or BrowserStateChangeDownloadStarted or BrowserStateChangeDownloadCompleted or BrowserStateChangeDownloadFailed`¶

- `BrowserStateChangeTabOpened object`¶

A tab this call's execution opened that remains open at its end —¶
the creation delta of the `tabs` inventory, not an event log.¶

Carries only the `tab_id`; the tab's `title` and `url` live on its¶
`tabs` entry, which must include the same `tab_id`. A tab opened¶
during a failed call gets no deferred `tab_opened`; it simply appears¶
in the next result's `tabs` inventory.¶

- `type: "tab_opened"`¶

- `tab_id: string`¶

The `tab_id` of the opened tab, present in `tabs`.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadStarted object`¶

A file download that started during this call.¶

- `type: "download_started"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadCompleted object`¶

A file download that finished during this call, reported with the¶
same `download_id` as its `download_started` — or without a prior¶
`download_started`, when the download finished during the call that¶
started it (at most one state change per `download_id` per result).¶

- `type: "download_completed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `path: optional string or null`¶

Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `size_bytes: optional number or null`¶

The completed download's size.¶

minimum: 0¶

- `BrowserStateChangeDownloadFailed object`¶

A file download that failed — or was cancelled — during this call.¶

- `type: "download_failed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `error: optional string or null`¶

The failure or cancellation detail, when known.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

### Browser State Change Download Completed¶

- `BrowserStateChangeDownloadCompleted object`¶

A file download that finished during this call, reported with the¶
same `download_id` as its `download_started` — or without a prior¶
`download_started`, when the download finished during the call that¶
started it (at most one state change per `download_id` per result).¶

- `type: "download_completed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `path: optional string or null`¶

Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `size_bytes: optional number or null`¶

The completed download's size.¶

minimum: 0¶

### Browser State Change Download Failed¶

- `BrowserStateChangeDownloadFailed object`¶

A file download that failed — or was cancelled — during this call.¶

- `type: "download_failed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `error: optional string or null`¶

The failure or cancellation detail, when known.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

### Browser State Change Download Started¶

- `BrowserStateChangeDownloadStarted object`¶

A file download that started during this call.¶

- `type: "download_started"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

### Browser State Change Tab Opened¶

- `BrowserStateChangeTabOpened object`¶

A tab this call's execution opened that remains open at its end —¶
the creation delta of the `tabs` inventory, not an event log.¶

Carries only the `tab_id`; the tab's `title` and `url` live on its¶
`tabs` entry, which must include the same `tab_id`. A tab opened¶
during a failed call gets no deferred `tab_opened`; it simply appears¶
in the next result's `tabs` inventory.¶

- `type: "tab_opened"`¶

- `tab_id: string`¶

The `tab_id` of the opened tab, present in `tabs`.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

### Browser State Tab Entry¶

- `BrowserStateTabEntry object`¶

One open browser tab reported in a `browser_state` block's `tabs`¶
inventory.¶

`tab_id` is the caller-assigned identifier for the tab; `title` and¶
`url` describe the page the tab is currently showing and may be empty¶
strings (a blank tab legitimately has both empty). `active` marks the¶
tab that is active after this call; whenever `tabs` is non-empty,¶
exactly one entry is marked.¶

- `tab_id: string`¶

The caller-assigned identifier for this tab, unique within the inventory.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `title: string`¶

The title of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The URL of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `active: optional boolean`¶

Whether this tab is the active tab after this call. Whenever `tabs` is non-empty, exactly one entry is marked `active: true`.¶

### Browser Switch Tab Config¶

- `BrowserSwitchTabConfig object`¶

`switch_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Toolset 20260801¶

- `BrowserToolset20260801 object`¶

The browser toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the browser tool family. The model is served¶
the family's tool with any members disabled via `configs` removed¶
from its schema.¶

- `type: "browser_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `configs: optional BrowserToolsetConfigs or null`¶

Per-member configuration for `browser_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional BrowserTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `close_tab: optional BrowserCloseTabConfig or null`¶

`close_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional BrowserDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `file_upload: optional BrowserFileUploadConfig or null`¶

`file_upload`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `find: optional BrowserFindConfig or null`¶

`find`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `form_input: optional BrowserFormInputConfig or null`¶

`form_input`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `get_page_text: optional BrowserGetPageTextConfig or null`¶

`get_page_text`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional BrowserHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hover: optional BrowserHoverConfig or null`¶

`hover`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `javascript_exec: optional BrowserJavascriptExecConfig or null`¶

`javascript_exec`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional BrowserKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional BrowserLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional BrowserLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional BrowserLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional BrowserLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `list_tabs: optional BrowserListTabsConfig or null`¶

`list_tabs`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional BrowserMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional BrowserMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `navigate: optional BrowserNavigateConfig or null`¶

`navigate`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `new_tab: optional BrowserNewTabConfig or null`¶

`new_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_console: optional BrowserReadConsoleConfig or null`¶

`read_console`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_network: optional BrowserReadNetworkConfig or null`¶

`read_network`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_page: optional BrowserReadPageConfig or null`¶

`read_page`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional BrowserRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional BrowserScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional BrowserScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll_to: optional BrowserScrollToConfig or null`¶

`scroll_to`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `switch_tab: optional BrowserSwitchTabConfig or null`¶

`switch_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional BrowserTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional BrowserWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional BrowserZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Toolset Configs¶

- `BrowserToolsetConfigs object`¶

Per-member configuration for `browser_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional BrowserTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `close_tab: optional BrowserCloseTabConfig or null`¶

`close_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional BrowserDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `file_upload: optional BrowserFileUploadConfig or null`¶

`file_upload`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `find: optional BrowserFindConfig or null`¶

`find`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `form_input: optional BrowserFormInputConfig or null`¶

`form_input`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `get_page_text: optional BrowserGetPageTextConfig or null`¶

`get_page_text`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional BrowserHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hover: optional BrowserHoverConfig or null`¶

`hover`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `javascript_exec: optional BrowserJavascriptExecConfig or null`¶

`javascript_exec`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional BrowserKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional BrowserLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional BrowserLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional BrowserLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional BrowserLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `list_tabs: optional BrowserListTabsConfig or null`¶

`list_tabs`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional BrowserMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional BrowserMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `navigate: optional BrowserNavigateConfig or null`¶

`navigate`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `new_tab: optional BrowserNewTabConfig or null`¶

`new_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_console: optional BrowserReadConsoleConfig or null`¶

`read_console`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_network: optional BrowserReadNetworkConfig or null`¶

`read_network`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_page: optional BrowserReadPageConfig or null`¶

`read_page`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional BrowserRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional BrowserScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional BrowserScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll_to: optional BrowserScrollToConfig or null`¶

`scroll_to`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `switch_tab: optional BrowserSwitchTabConfig or null`¶

`switch_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional BrowserTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional BrowserWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional BrowserZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Triple Click Config¶

- `BrowserTripleClickConfig object`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Type Config¶

- `BrowserTypeConfig object`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Wait Config¶

- `BrowserWaitConfig object`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Browser Zoom Config¶

- `BrowserZoomConfig object`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Cache Control Ephemeral¶

- `CacheControlEphemeral object`¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

### Cache Creation¶

- `CacheCreation object`¶

- `ephemeral_1h_input_tokens: number`¶

The number of input tokens used to create the 1 hour cache entry.¶

default: 0, minimum: 0¶

- `ephemeral_5m_input_tokens: number`¶

The number of input tokens used to create the 5 minute cache entry.¶

default: 0, minimum: 0¶

### Citation Char Location¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

### Citation Char Location Param¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

### Citation Content Block Location¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

### Citation Content Block Location Param¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

### Citation Page Location¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

### Citation Page Location Param¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

### Citation Search Result Location Param¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

### Citation Web Search Result Location Param¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

### Citations Config¶

- `CitationsConfig object`¶

- `enabled: boolean`¶

default: false¶

### Citations Config Param¶

- `CitationsConfigParam object`¶

- `enabled: optional boolean`¶

### Citations Delta¶

- `CitationsDelta object`¶

- `type: "citations_delta"`¶

default: citations_delta¶

- `citation: CitationCharLocation or CitationPageLocation or CitationContentBlockLocation or 2 more`¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

### Citations Search Result Location¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

### Citations Web Search Result Location¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

### Code Execution Output Block¶

- `CodeExecutionOutputBlock object`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

### Code Execution Output Block Param¶

- `CodeExecutionOutputBlockParam object`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

### Code Execution Result Block¶

- `CodeExecutionResultBlock object`¶

- `type: "code_execution_result"`¶

default: code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

### Code Execution Result Block Param¶

- `CodeExecutionResultBlockParam object`¶

- `type: "code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

### Code Execution Tool 20250522¶

- `CodeExecutionTool20250522 object`¶

- `type: "code_execution_20250522"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Code Execution Tool 20250825¶

- `CodeExecutionTool20250825 object`¶

- `type: "code_execution_20250825"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Code Execution Tool 20260120¶

- `CodeExecutionTool20260120 object`¶

Code execution tool with REPL state persistence (daemon mode + gVisor checkpoint).¶

- `type: "code_execution_20260120"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Code Execution Tool 20260521¶

- `CodeExecutionTool20260521 object`¶

Code execution tool with REPL state persistence.¶

- `type: "code_execution_20260521"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Code Execution Tool Result Block¶

- `CodeExecutionToolResultBlock object`¶

- `type: "code_execution_tool_result"`¶

default: code_execution_tool_result¶

- `content: CodeExecutionToolResultBlockContent`¶

- `CodeExecutionToolResultError object`¶

- `type: "code_execution_tool_result_error"`¶

default: code_execution_tool_result_error¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlock object`¶

- `type: "code_execution_result"`¶

default: code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlock object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

default: encrypted_code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Code Execution Tool Result Block Content¶

- `CodeExecutionToolResultBlockContent = CodeExecutionToolResultError or CodeExecutionResultBlock or EncryptedCodeExecutionResultBlock`¶

- `CodeExecutionToolResultError object`¶

- `type: "code_execution_tool_result_error"`¶

default: code_execution_tool_result_error¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlock object`¶

- `type: "code_execution_result"`¶

default: code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlock object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

default: encrypted_code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

### Code Execution Tool Result Block Param¶

- `CodeExecutionToolResultBlockParam object`¶

- `type: "code_execution_tool_result"`¶

- `content: CodeExecutionToolResultBlockParamContent`¶

- `CodeExecutionToolResultErrorParam object`¶

- `type: "code_execution_tool_result_error"`¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlockParam object`¶

- `type: "code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlockParam object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

### Code Execution Tool Result Block Param Content¶

- `CodeExecutionToolResultBlockParamContent = CodeExecutionToolResultErrorParam or CodeExecutionResultBlockParam or EncryptedCodeExecutionResultBlockParam`¶

- `CodeExecutionToolResultErrorParam object`¶

- `type: "code_execution_tool_result_error"`¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlockParam object`¶

- `type: "code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlockParam object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

### Code Execution Tool Result Error¶

- `CodeExecutionToolResultError object`¶

- `type: "code_execution_tool_result_error"`¶

default: code_execution_tool_result_error¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

### Code Execution Tool Result Error Code¶

- `CodeExecutionToolResultErrorCode = "invalid_tool_input" or "unavailable" or "too_many_requests" or "execution_time_exceeded"`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

### Code Execution Tool Result Error Param¶

- `CodeExecutionToolResultErrorParam object`¶

- `type: "code_execution_tool_result_error"`¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

### Computer Cursor Position Config¶

- `ComputerCursorPositionConfig object`¶

`cursor_position`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Double Click Config¶

- `ComputerDoubleClickConfig object`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Hold Key Config¶

- `ComputerHoldKeyConfig object`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Key Config¶

- `ComputerKeyConfig object`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Left Click Config¶

- `ComputerLeftClickConfig object`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Left Click Drag Config¶

- `ComputerLeftClickDragConfig object`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Left Mouse Down Config¶

- `ComputerLeftMouseDownConfig object`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Left Mouse Up Config¶

- `ComputerLeftMouseUpConfig object`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Middle Click Config¶

- `ComputerMiddleClickConfig object`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Mouse Move Config¶

- `ComputerMouseMoveConfig object`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Right Click Config¶

- `ComputerRightClickConfig object`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Screenshot Config¶

- `ComputerScreenshotConfig object`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Scroll Config¶

- `ComputerScrollConfig object`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Toolset 20260801¶

- `ComputerToolset20260801 object`¶

The computer toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the computer tool family. The model is¶
served the family's tool with any members disabled via `configs`¶
removed from its schema. Every member is enabled by default, zoom¶
included. The single-tool options `display_number` and¶
`enable_zoom` are not fields of a toolset entry — it carries only¶
`type`, `configs`, and `cache_control`; zoom is controlled¶
via `configs.zoom.enabled`.¶

- `type: "computer_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `configs: optional ComputerToolsetConfigs or null`¶

Per-member configuration for `computer_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional ComputerTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `cursor_position: optional ComputerCursorPositionConfig or null`¶

`cursor_position`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional ComputerDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional ComputerHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional ComputerKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional ComputerLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional ComputerLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional ComputerLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional ComputerLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional ComputerMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional ComputerMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional ComputerRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional ComputerScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional ComputerScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional ComputerTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional ComputerWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional ComputerZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Toolset Configs¶

- `ComputerToolsetConfigs object`¶

Per-member configuration for `computer_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional ComputerTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `cursor_position: optional ComputerCursorPositionConfig or null`¶

`cursor_position`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional ComputerDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional ComputerHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional ComputerKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional ComputerLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional ComputerLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional ComputerLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional ComputerLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional ComputerMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional ComputerMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional ComputerRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional ComputerScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional ComputerScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional ComputerTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional ComputerWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional ComputerZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Triple Click Config¶

- `ComputerTripleClickConfig object`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Type Config¶

- `ComputerTypeConfig object`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Wait Config¶

- `ComputerWaitConfig object`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Computer Zoom Config¶

- `ComputerZoomConfig object`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

### Container¶

- `Container object`¶

Information about the container used in the request (for the code execution tool)¶

- `id: string`¶

Identifier for the container used in this request¶

- `expires_at: string`¶

The time at which the container will expire.¶

format: date-time¶

- `skills: array of ContainerSkill or null`¶

Skills loaded in the container¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: string`¶

The resolved version: a skill version ID for custom skills.¶

maxLength: 64, minLength: 1¶

### Container Params¶

- `ContainerParams object`¶

Container parameters with skills to be loaded.¶

- `id: optional string or null`¶

Container id¶

- `skills: optional array of SkillParams or null`¶

List of skills to load in the container¶

maxItems: 20¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: optional string`¶

Skill version or 'latest' for most recent version¶

maxLength: 64, minLength: 1¶

### Container Skill¶

- `ContainerSkill object`¶

A skill that was loaded in a container (response model).¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: string`¶

The resolved version: a skill version ID for custom skills.¶

maxLength: 64, minLength: 1¶

### Container Upload Block¶

- `ContainerUploadBlock object`¶

Response model for a file uploaded to the container.¶

- `type: "container_upload"`¶

default: container_upload¶

- `file_id: string`¶

### Container Upload Block Param¶

- `ContainerUploadBlockParam object`¶

A content block that represents a file to be uploaded to the container¶
Files uploaded via this block will be available in the container's input directory.¶

- `type: "container_upload"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

### Content Block¶

- `ContentBlock = TextBlock or ThinkingBlock or RedactedThinkingBlock or 9 more`¶

- `TextBlock object`¶

- `type: "text"`¶

default: text¶

- `citations: array of TextCitation or null`¶

Citations supporting the text block.¶

The type of citation returned will depend on the type of document being cited. Citing a PDF results in `page_location`, plain text results in `char_location`, and content document results in `content_block_location`.¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `text: string`¶

minLength: 0¶

- `ThinkingBlock object`¶

- `type: "thinking"`¶

default: thinking¶

- `signature: string`¶

A value used to verify that this thinking block was generated by Claude when it is passed back to the API.¶

This is an opaque field and should not be interpreted or parsed. When passing thinking blocks back to the API (required when using tools with extended thinking), pass them back exactly as received, with this field intact.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `thinking: string`¶

The text of Claude's thinking process for this block.¶

- `RedactedThinkingBlock object`¶

- `type: "redacted_thinking"`¶

default: redacted_thinking¶

- `data: string`¶

The contents of this redacted thinking block, returned when portions of the model's thinking were safety-redacted. This field is opaque and encrypted, with no readable content.¶

Pass `redacted_thinking` blocks back to the API unchanged when continuing a multi-turn conversation.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking#redacted-thinking-blocks) for details.¶

- `ToolUseBlock object`¶

- `type: "tool_use"`¶

default: tool_use¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: string`¶

minLength: 1¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlock object`¶

- `type: "server_tool_use"`¶

default: server_tool_use¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `WebSearchToolResultBlock object`¶

- `type: "web_search_tool_result"`¶

default: web_search_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebSearchToolResultBlockContent`¶

- `WebSearchToolResultError object`¶

- `type: "web_search_tool_result_error"`¶

default: web_search_tool_result_error¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `array of WebSearchResultBlock`¶

- `type: "web_search_result"`¶

default: web_search_result¶

- `encrypted_content: string`¶

- `page_age: string or null`¶

- `title: string`¶

- `url: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `WebFetchToolResultBlock object`¶

- `type: "web_fetch_tool_result"`¶

default: web_fetch_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebFetchToolResultErrorBlock or WebFetchBlock`¶

- `WebFetchToolResultErrorBlock object`¶

- `type: "web_fetch_tool_result_error"`¶

default: web_fetch_tool_result_error¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlock object`¶

- `type: "web_fetch_result"`¶

default: web_fetch_result¶

- `content: DocumentBlock`¶

- `type: "document"`¶

default: document¶

- `citations: CitationsConfig or null`¶

Citation configuration for the document¶

- `enabled: boolean`¶

default: false¶

- `source: Base64PDFSource or PlainTextSource`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `title: string or null`¶

The title of the document¶

- `retrieved_at: string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `url: string`¶

Fetched content URL¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `CodeExecutionToolResultBlock object`¶

- `type: "code_execution_tool_result"`¶

default: code_execution_tool_result¶

- `content: CodeExecutionToolResultBlockContent`¶

- `CodeExecutionToolResultError object`¶

- `type: "code_execution_tool_result_error"`¶

default: code_execution_tool_result_error¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlock object`¶

- `type: "code_execution_result"`¶

default: code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlock object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

default: encrypted_code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `BashCodeExecutionToolResultBlock object`¶

- `type: "bash_code_execution_tool_result"`¶

default: bash_code_execution_tool_result¶

- `content: BashCodeExecutionToolResultError or BashCodeExecutionResultBlock`¶

- `BashCodeExecutionToolResultError object`¶

- `type: "bash_code_execution_tool_result_error"`¶

default: bash_code_execution_tool_result_error¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlock object`¶

- `type: "bash_code_execution_result"`¶

default: bash_code_execution_result¶

- `content: array of BashCodeExecutionOutputBlock`¶

- `type: "bash_code_execution_output"`¶

default: bash_code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `TextEditorCodeExecutionToolResultBlock object`¶

- `type: "text_editor_code_execution_tool_result"`¶

default: text_editor_code_execution_tool_result¶

- `content: TextEditorCodeExecutionToolResultError or TextEditorCodeExecutionViewResultBlock or TextEditorCodeExecutionCreateResultBlock or TextEditorCodeExecutionStrReplaceResultBlock`¶

- `TextEditorCodeExecutionToolResultError object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

default: text_editor_code_execution_tool_result_error¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: string or null`¶

- `TextEditorCodeExecutionViewResultBlock object`¶

- `type: "text_editor_code_execution_view_result"`¶

default: text_editor_code_execution_view_result¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: number or null`¶

- `start_line: number or null`¶

- `total_lines: number or null`¶

- `TextEditorCodeExecutionCreateResultBlock object`¶

- `type: "text_editor_code_execution_create_result"`¶

default: text_editor_code_execution_create_result¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlock object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

default: text_editor_code_execution_str_replace_result¶

- `lines: array of string or null`¶

- `new_lines: number or null`¶

- `new_start: number or null`¶

- `old_lines: number or null`¶

- `old_start: number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ToolSearchToolResultBlock object`¶

- `type: "tool_search_tool_result"`¶

default: tool_search_tool_result¶

- `content: ToolSearchToolResultError or ToolSearchToolSearchResultBlock`¶

- `ToolSearchToolResultError object`¶

- `type: "tool_search_tool_result_error"`¶

default: tool_search_tool_result_error¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: string or null`¶

- `ToolSearchToolSearchResultBlock object`¶

- `type: "tool_search_tool_search_result"`¶

default: tool_search_tool_search_result¶

- `tool_references: array of ToolReferenceBlock`¶

- `type: "tool_reference"`¶

default: tool_reference¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ContainerUploadBlock object`¶

Response model for a file uploaded to the container.¶

- `type: "container_upload"`¶

default: container_upload¶

- `file_id: string`¶

### Content Block Param¶

- `ContentBlockParam = TextBlockParam or ImageBlockParam or DocumentBlockParam or 13 more`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

- `DocumentBlockParam object`¶

- `type: "document"`¶

- `source: Base64PDFSource or PlainTextSource or ContentBlockSource or 2 more`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `ContentBlockSource object`¶

- `type: "content"`¶

- `content: string or array of ContentBlockSourceContent`¶

- `string`¶

- `ContentBlockSourceContent = array of ContentBlockSourceContent`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `URLPDFSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileDocumentSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

- `enabled: optional boolean`¶

- `context: optional string or null`¶

minLength: 1¶

- `title: optional string or null`¶

maxLength: 500, minLength: 1¶

- `SearchResultBlockParam object`¶

- `type: "search_result"`¶

- `content: array of TextBlockParam`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional array of TextCitationParam or null`¶

- `source: string`¶

- `title: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam`¶

- `ThinkingBlockParam object`¶

- `type: "thinking"`¶

- `signature: string`¶

The `signature` value of this thinking block, exactly as returned by the API in a previous response. Used to verify that the block was generated by Claude.¶

Thinking blocks must be passed back unmodified and in their original order; a modified block results in a 400 `invalid_request_error`.¶

- `thinking: string`¶

The `thinking` text of this block as returned by the API.¶

- `RedactedThinkingBlockParam object`¶

- `type: "redacted_thinking"`¶

- `data: string`¶

The `data` value of this redacted thinking block, exactly as returned by the API in a previous response. Opaque and encrypted; pass it back unchanged.¶

- `ToolUseBlockParam object`¶

- `type: "tool_use"`¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `input: map[unknown]`¶

- `name: string`¶

maxLength: 200, minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family this member belongs to.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ToolResultBlockParam object`¶

- `type: "tool_result"`¶

- `tool_use_id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `content: optional string or array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `string`¶

- `array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `SearchResultBlockParam object`¶

- `DocumentBlockParam object`¶

- `ToolReferenceBlockParam object`¶

Tool reference block that can be included in tool_result content.¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BrowserStateBlockParam object`¶

The caller's browser state after a browser toolset member call —¶
the full inventory of open tabs, which tab is active, and any side¶
effects (tabs opened, download state changes) the call produced.¶

At most one per `tool_result`, only on a non-error result answering a¶
browser toolset member `tool_use`. The server renders the¶
model-visible text from it; the model never sees the raw fields.¶

- `type: "browser_state"`¶

- `tabs: array of BrowserStateTabEntry`¶

All tabs open in the browser after this call — the full inventory, not a delta. May be empty. Whenever non-empty, exactly one entry carries `active: true`.¶

maxItems: 100¶

- `tab_id: string`¶

The caller-assigned identifier for this tab, unique within the inventory.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `title: string`¶

The title of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The URL of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `active: optional boolean`¶

Whether this tab is the active tab after this call. Whenever `tabs` is non-empty, exactly one entry is marked `active: true`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `state_changes: optional array of BrowserStateChange or null`¶

Tabs opened and download state changes during this call. "Nothing to report" is expressed by omitting the field, never by an empty list.¶

maxItems: 200, minItems: 1¶

- `BrowserStateChangeTabOpened object`¶

A tab this call's execution opened that remains open at its end —¶
the creation delta of the `tabs` inventory, not an event log.¶

Carries only the `tab_id`; the tab's `title` and `url` live on its¶
`tabs` entry, which must include the same `tab_id`. A tab opened¶
during a failed call gets no deferred `tab_opened`; it simply appears¶
in the next result's `tabs` inventory.¶

- `type: "tab_opened"`¶

- `tab_id: string`¶

The `tab_id` of the opened tab, present in `tabs`.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadStarted object`¶

A file download that started during this call.¶

- `type: "download_started"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadCompleted object`¶

A file download that finished during this call, reported with the¶
same `download_id` as its `download_started` — or without a prior¶
`download_started`, when the download finished during the call that¶
started it (at most one state change per `download_id` per result).¶

- `type: "download_completed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `path: optional string or null`¶

Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `size_bytes: optional number or null`¶

The completed download's size.¶

minimum: 0¶

- `BrowserStateChangeDownloadFailed object`¶

A file download that failed — or was cancelled — during this call.¶

- `type: "download_failed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `error: optional string or null`¶

The failure or cancellation detail, when known.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `is_error: optional boolean`¶

- `toolset_name: optional string or null`¶

For a toolset member tool_result, the toolset family of the paired tool_use.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlockParam object`¶

- `type: "server_tool_use"`¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `WebSearchToolResultBlockParam object`¶

- `type: "web_search_tool_result"`¶

- `content: WebSearchToolResultBlockParamContent`¶

- `WebSearchToolResultBlockItem = array of WebSearchResultBlockParam`¶

- `type: "web_search_result"`¶

- `encrypted_content: string`¶

- `title: string`¶

- `url: string`¶

- `page_age: optional string or null`¶

- `WebSearchToolRequestError object`¶

- `type: "web_search_tool_result_error"`¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `WebFetchToolResultBlockParam object`¶

- `type: "web_fetch_tool_result"`¶

- `content: WebFetchToolResultErrorBlockParam or WebFetchBlockParam`¶

- `WebFetchToolResultErrorBlockParam object`¶

- `type: "web_fetch_tool_result_error"`¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlockParam object`¶

- `type: "web_fetch_result"`¶

- `content: DocumentBlockParam`¶

- `url: string`¶

Fetched content URL¶

- `retrieved_at: optional string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `CodeExecutionToolResultBlockParam object`¶

- `type: "code_execution_tool_result"`¶

- `content: CodeExecutionToolResultBlockParamContent`¶

- `CodeExecutionToolResultErrorParam object`¶

- `type: "code_execution_tool_result_error"`¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlockParam object`¶

- `type: "code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlockParam object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BashCodeExecutionToolResultBlockParam object`¶

- `type: "bash_code_execution_tool_result"`¶

- `content: BashCodeExecutionToolResultErrorParam or BashCodeExecutionResultBlockParam`¶

- `BashCodeExecutionToolResultErrorParam object`¶

- `type: "bash_code_execution_tool_result_error"`¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlockParam object`¶

- `type: "bash_code_execution_result"`¶

- `content: array of BashCodeExecutionOutputBlockParam`¶

- `type: "bash_code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `TextEditorCodeExecutionToolResultBlockParam object`¶

- `type: "text_editor_code_execution_tool_result"`¶

- `content: TextEditorCodeExecutionToolResultErrorParam or TextEditorCodeExecutionViewResultBlockParam or TextEditorCodeExecutionCreateResultBlockParam or TextEditorCodeExecutionStrReplaceResultBlockParam`¶

- `TextEditorCodeExecutionToolResultErrorParam object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: optional string or null`¶

- `TextEditorCodeExecutionViewResultBlockParam object`¶

- `type: "text_editor_code_execution_view_result"`¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: optional number or null`¶

- `start_line: optional number or null`¶

- `total_lines: optional number or null`¶

- `TextEditorCodeExecutionCreateResultBlockParam object`¶

- `type: "text_editor_code_execution_create_result"`¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlockParam object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

- `lines: optional array of string or null`¶

- `new_lines: optional number or null`¶

- `new_start: optional number or null`¶

- `old_lines: optional number or null`¶

- `old_start: optional number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `ToolSearchToolResultBlockParam object`¶

- `type: "tool_search_tool_result"`¶

- `content: ToolSearchToolResultErrorParam or ToolSearchToolSearchResultBlockParam`¶

- `ToolSearchToolResultErrorParam object`¶

- `type: "tool_search_tool_result_error"`¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: optional string or null`¶

- `ToolSearchToolSearchResultBlockParam object`¶

- `type: "tool_search_tool_search_result"`¶

- `tool_references: array of ToolReferenceBlockParam`¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `ContainerUploadBlockParam object`¶

A content block that represents a file to be uploaded to the container¶
Files uploaded via this block will be available in the container's input directory.¶

- `type: "container_upload"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

### Content Block Source¶

- `ContentBlockSource object`¶

- `type: "content"`¶

- `content: string or array of ContentBlockSourceContent`¶

- `string`¶

- `ContentBlockSourceContent = array of ContentBlockSourceContent`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

### Content Block Source Content¶

- `ContentBlockSourceContent = TextBlockParam or ImageBlockParam`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

### Direct Caller¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

### Document Block¶

- `DocumentBlock object`¶

- `type: "document"`¶

default: document¶

- `citations: CitationsConfig or null`¶

Citation configuration for the document¶

- `enabled: boolean`¶

default: false¶

- `source: Base64PDFSource or PlainTextSource`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `title: string or null`¶

The title of the document¶

### Document Block Param¶

- `DocumentBlockParam object`¶

- `type: "document"`¶

- `source: Base64PDFSource or PlainTextSource or ContentBlockSource or 2 more`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `ContentBlockSource object`¶

- `type: "content"`¶

- `content: string or array of ContentBlockSourceContent`¶

- `string`¶

- `ContentBlockSourceContent = array of ContentBlockSourceContent`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

- `URLPDFSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileDocumentSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

- `enabled: optional boolean`¶

- `context: optional string or null`¶

minLength: 1¶

- `title: optional string or null`¶

maxLength: 500, minLength: 1¶

### Encrypted Code Execution Result Block¶

- `EncryptedCodeExecutionResultBlock object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

default: encrypted_code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

### Encrypted Code Execution Result Block Param¶

- `EncryptedCodeExecutionResultBlockParam object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

### File Document Source¶

- `FileDocumentSource object`¶

- `type: "file"`¶

- `file_id: string`¶

### File Image Source¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

### Image Block Param¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

### Image Transformations Param¶

- `ImageTransformationsParam object`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

### Input JSON Delta¶

- `InputJSONDelta object`¶

- `type: "input_json_delta"`¶

default: input_json_delta¶

- `partial_json: string`¶

### JSON Output Format¶

- `JSONOutputFormat object`¶

- `type: "json_schema"`¶

- `schema: map[unknown]`¶

The JSON schema of the format¶

### Memory Tool 20250818¶

- `MemoryTool20250818 object`¶

- `type: "memory_20250818"`¶

- `name: "memory"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Message¶

- `Message object`¶

- `type: "message"`¶

Object type.¶

For Messages, this is always `"message"`.¶

default: message¶

- `id: string`¶

Unique object identifier.¶

The format and length of IDs may change over time.¶

- `container: Container or null`¶

Information about the container used in the request (for the code execution tool)¶

- `id: string`¶

Identifier for the container used in this request¶

- `expires_at: string`¶

The time at which the container will expire.¶

format: date-time¶

- `skills: array of ContainerSkill or null`¶

Skills loaded in the container¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: string`¶

The resolved version: a skill version ID for custom skills.¶

maxLength: 64, minLength: 1¶

- `content: array of ContentBlock`¶

Content generated by the model.¶

This is an array of content blocks, each of which has a `type` that determines its shape.¶

Example:¶

```json¶
[{"type": "text", "text": "Hi, I'm Claude."}]¶
```¶

If the request input `messages` ended with an `assistant` turn, then the response `content` will continue directly from that last turn. You can use this to constrain the model's output.¶

For example, if the input `messages` were:¶

```json¶
[¶
{"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},¶
{"role": "assistant", "content": "The best answer is ("}¶
]¶
```¶

Then the response `content` might be:¶

```json¶
[{"type": "text", "text": "B)"}]¶
```¶

- `TextBlock object`¶

- `type: "text"`¶

default: text¶

- `citations: array of TextCitation or null`¶

Citations supporting the text block.¶

The type of citation returned will depend on the type of document being cited. Citing a PDF results in `page_location`, plain text results in `char_location`, and content document results in `content_block_location`.¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `text: string`¶

minLength: 0¶

- `ThinkingBlock object`¶

- `type: "thinking"`¶

default: thinking¶

- `signature: string`¶

A value used to verify that this thinking block was generated by Claude when it is passed back to the API.¶

This is an opaque field and should not be interpreted or parsed. When passing thinking blocks back to the API (required when using tools with extended thinking), pass them back exactly as received, with this field intact.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `thinking: string`¶

The text of Claude's thinking process for this block.¶

- `RedactedThinkingBlock object`¶

- `type: "redacted_thinking"`¶

default: redacted_thinking¶

- `data: string`¶

The contents of this redacted thinking block, returned when portions of the model's thinking were safety-redacted. This field is opaque and encrypted, with no readable content.¶

Pass `redacted_thinking` blocks back to the API unchanged when continuing a multi-turn conversation.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking#redacted-thinking-blocks) for details.¶

- `ToolUseBlock object`¶

- `type: "tool_use"`¶

default: tool_use¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: string`¶

minLength: 1¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlock object`¶

- `type: "server_tool_use"`¶

default: server_tool_use¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `WebSearchToolResultBlock object`¶

- `type: "web_search_tool_result"`¶

default: web_search_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebSearchToolResultBlockContent`¶

- `WebSearchToolResultError object`¶

- `type: "web_search_tool_result_error"`¶

default: web_search_tool_result_error¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `array of WebSearchResultBlock`¶

- `type: "web_search_result"`¶

default: web_search_result¶

- `encrypted_content: string`¶

- `page_age: string or null`¶

- `title: string`¶

- `url: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `WebFetchToolResultBlock object`¶

- `type: "web_fetch_tool_result"`¶

default: web_fetch_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebFetchToolResultErrorBlock or WebFetchBlock`¶

- `WebFetchToolResultErrorBlock object`¶

- `type: "web_fetch_tool_result_error"`¶

default: web_fetch_tool_result_error¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlock object`¶

- `type: "web_fetch_result"`¶

default: web_fetch_result¶

- `content: DocumentBlock`¶

- `type: "document"`¶

default: document¶

- `citations: CitationsConfig or null`¶

Citation configuration for the document¶

- `enabled: boolean`¶

default: false¶

- `source: Base64PDFSource or PlainTextSource`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `title: string or null`¶

The title of the document¶

- `retrieved_at: string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `url: string`¶

Fetched content URL¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `CodeExecutionToolResultBlock object`¶

- `type: "code_execution_tool_result"`¶

default: code_execution_tool_result¶

- `content: CodeExecutionToolResultBlockContent`¶

- `CodeExecutionToolResultError object`¶

- `type: "code_execution_tool_result_error"`¶

default: code_execution_tool_result_error¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlock object`¶

- `type: "code_execution_result"`¶

default: code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlock object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

default: encrypted_code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `BashCodeExecutionToolResultBlock object`¶

- `type: "bash_code_execution_tool_result"`¶

default: bash_code_execution_tool_result¶

- `content: BashCodeExecutionToolResultError or BashCodeExecutionResultBlock`¶

- `BashCodeExecutionToolResultError object`¶

- `type: "bash_code_execution_tool_result_error"`¶

default: bash_code_execution_tool_result_error¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlock object`¶

- `type: "bash_code_execution_result"`¶

default: bash_code_execution_result¶

- `content: array of BashCodeExecutionOutputBlock`¶

- `type: "bash_code_execution_output"`¶

default: bash_code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `TextEditorCodeExecutionToolResultBlock object`¶

- `type: "text_editor_code_execution_tool_result"`¶

default: text_editor_code_execution_tool_result¶

- `content: TextEditorCodeExecutionToolResultError or TextEditorCodeExecutionViewResultBlock or TextEditorCodeExecutionCreateResultBlock or TextEditorCodeExecutionStrReplaceResultBlock`¶

- `TextEditorCodeExecutionToolResultError object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

default: text_editor_code_execution_tool_result_error¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: string or null`¶

- `TextEditorCodeExecutionViewResultBlock object`¶

- `type: "text_editor_code_execution_view_result"`¶

default: text_editor_code_execution_view_result¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: number or null`¶

- `start_line: number or null`¶

- `total_lines: number or null`¶

- `TextEditorCodeExecutionCreateResultBlock object`¶

- `type: "text_editor_code_execution_create_result"`¶

default: text_editor_code_execution_create_result¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlock object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

default: text_editor_code_execution_str_replace_result¶

- `lines: array of string or null`¶

- `new_lines: number or null`¶

- `new_start: number or null`¶

- `old_lines: number or null`¶

- `old_start: number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ToolSearchToolResultBlock object`¶

- `type: "tool_search_tool_result"`¶

default: tool_search_tool_result¶

- `content: ToolSearchToolResultError or ToolSearchToolSearchResultBlock`¶

- `ToolSearchToolResultError object`¶

- `type: "tool_search_tool_result_error"`¶

default: tool_search_tool_result_error¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: string or null`¶

- `ToolSearchToolSearchResultBlock object`¶

- `type: "tool_search_tool_search_result"`¶

default: tool_search_tool_search_result¶

- `tool_references: array of ToolReferenceBlock`¶

- `type: "tool_reference"`¶

default: tool_reference¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ContainerUploadBlock object`¶

Response model for a file uploaded to the container.¶

- `type: "container_upload"`¶

default: container_upload¶

- `file_id: string`¶

- `model: Model`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1"`¶

Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows¶

- `"claude-mythos-5-1"`¶

Our most capable model for cybersecurity and biology research, available through trusted access programs¶

- `"claude-sonnet-5"`¶

High-performance model for coding and agents¶

- `"claude-fable-5"`¶

Next generation of intelligence for the hardest knowledge work and coding problems¶

- `"claude-mythos-5"`¶

Most capable model for cybersecurity and biology research¶

- `"claude-opus-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-8"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-7"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-mythos-preview"`¶

New class of intelligence, strongest in coding and cybersecurity¶

- `"claude-opus-4-6"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-6"`¶

Best combination of speed and intelligence¶

- `"claude-haiku-4-5"`¶

Fastest model with near-frontier intelligence¶

- `"claude-haiku-4-5-20251001"`¶

Fastest model with near-frontier intelligence¶

- `"claude-opus-4-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-5-20251101"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-5"`¶

High-performance model for agents and coding¶

- `"claude-sonnet-4-5-20250929"`¶

High-performance model for agents and coding¶

- `string`¶

- `role: "assistant"`¶

Conversational role of the generated message.¶

This will always be `"assistant"`.¶

default: assistant¶

- `stop_details: RefusalStopDetails or null`¶

Structured information about a refusal.¶

- `type: "refusal"`¶

default: refusal¶

- `category: "cyber" or "bio" or "frontier_llm" or 2 more or null`¶

The policy category that triggered a refusal.¶

- `"cyber"`¶

The request could enable cyber harm, such as malware or exploit development. Benign cybersecurity work can also trigger this category.¶

- `"bio"`¶

The request could enable biological harm, such as dangerous lab methods. Beneficial life sciences work can also trigger this category.¶

- `"frontier_llm"`¶

The request could assist the development of competing AI models, which is restricted under [Anthropic's commercial terms](https://www.anthropic.com/legal/commercial-terms). Benign machine learning work can also trigger this category.¶

- `"reasoning_extraction"`¶

The request asks the model to reproduce its internal reasoning in the response text. To get reasoning in a structured form instead, use [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/adaptive-thinking).¶

- `"general_harms"`¶

The request could be related to an area that was determined as harmful. Benign work might sometimes trigger this category.¶

- `explanation: string or null`¶

Human-readable explanation of the refusal.¶

This text is not guaranteed to be stable. `null` when no explanation is available for the category.¶

- `stop_reason: StopReason or null`¶

The reason that we stopped.¶

This may be one the following values:¶

* `"end_turn"`: the model reached a natural stopping point¶
* `"max_tokens"`: we exceeded the requested `max_tokens` or the model's maximum¶
* `"stop_sequence"`: one of your provided custom `stop_sequences` was generated¶
* `"tool_use"`: the model invoked one or more tools¶
* `"pause_turn"`: we paused a long-running turn. You may provide the response back as-is in a subsequent request to let the model continue.¶
* `"refusal"`: when streaming classifiers intervene to handle potential policy violations¶
* `"model_context_window_exceeded"`: we exceeded the model's context window¶

In non-streaming mode this value is always non-null. In streaming mode, it is null in the `message_start` event and non-null otherwise.¶

- `"end_turn"`¶

- `"max_tokens"`¶

- `"stop_sequence"`¶

- `"tool_use"`¶

- `"pause_turn"`¶

- `"refusal"`¶

- `"model_context_window_exceeded"`¶

- `stop_sequence: string or null`¶

Which custom stop sequence was generated, if any.¶

This value will be a non-null string if one of your custom stop sequences was generated.¶

- `usage: Usage`¶

Billing and rate-limit usage.¶

Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.¶

Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.¶

For example, `output_tokens` will be non-zero, even for an empty string response from Claude.¶

Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.¶

- `cache_creation: CacheCreation or null`¶

Breakdown of cached tokens by TTL¶

- `ephemeral_1h_input_tokens: number`¶

The number of input tokens used to create the 1 hour cache entry.¶

default: 0, minimum: 0¶

- `ephemeral_5m_input_tokens: number`¶

The number of input tokens used to create the 5 minute cache entry.¶

default: 0, minimum: 0¶

- `cache_creation_input_tokens: number or null`¶

The number of input tokens used to create the cache entry.¶

minimum: 0¶

- `cache_read_input_tokens: number or null`¶

The number of input tokens read from the cache.¶

minimum: 0¶

- `inference_geo: string or null`¶

The geographic region where inference was performed for this request.¶

- `input_tokens: number`¶

The number of input tokens which were used.¶

minimum: 0¶

- `output_tokens: number`¶

The number of output tokens which were used.¶

minimum: 0¶

- `output_tokens_details: OutputTokensDetails or null`¶

Breakdown of output tokens by category.¶

`output_tokens` remains the inclusive, authoritative total used for billing.¶
This object provides a read-only decomposition for observability — for example,¶
how many of the billed output tokens were spent on internal reasoning that may¶
have been summarized before being returned to you.¶

- `thinking_tokens: number`¶

Number of output tokens the model generated as internal reasoning, including¶
the thinking-block delimiter tokens.¶

Reflects the raw reasoning the model produced, not the (possibly shorter)¶
summarized thinking text returned in the response body. Computed by¶
re-tokenizing the raw reasoning text, so it may differ from the model's exact¶
generation count by a small number of tokens. Always ≤ `output_tokens`;¶
`output_tokens - thinking_tokens` approximates the non-reasoning output.¶

default: 0, minimum: 0¶

- `server_tool_use: ServerToolUsage or null`¶

The number of server tool requests.¶

- `web_fetch_requests: number`¶

The number of web fetch tool requests.¶

default: 0, minimum: 0¶

- `web_search_requests: number`¶

The number of web search tool requests.¶

default: 0, minimum: 0¶

- `service_tier: "standard" or "priority" or "batch" or null`¶

If the request used the priority, standard, or batch tier.¶

- `"standard"`¶

- `"priority"`¶

- `"batch"`¶

### Message Count Tokens Tool¶

- `MessageCountTokensTool = Tool or ToolBash20250124 or CodeExecutionTool20250522 or 18 more`¶

- `Tool object`¶

- `type: optional "custom" or null`¶

- `input_schema: object`¶

[JSON schema](https://json-schema.org/draft/2020-12) for this tool's input.¶

This defines the shape of the `input` that your tool accepts and that the model will produce.¶

- `type: "object"`¶

- `properties: optional map[unknown] or null`¶

- `required: optional array of string or null`¶

- `name: string`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

maxLength: 128, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,128}$¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `description: optional string`¶

Description of what this tool does.¶

Tool descriptions should be as detailed as possible. The more information that the model has about what the tool is and how to use it, the better it will perform. You can use natural language descriptions to reinforce important aspects of the tool input JSON schema.¶

- `eager_input_streaming: optional boolean or null`¶

Enable eager input streaming for this tool. When true, tool input parameters will be streamed incrementally as they are generated, and types will be inferred on-the-fly rather than buffering the full JSON output. When false, streaming is disabled for this tool even if the fine-grained-tool-streaming beta is active. When null (default), uses the default behavior based on beta headers.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolBash20250124 object`¶

- `type: "bash_20250124"`¶

- `name: "bash"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20250522 object`¶

- `type: "code_execution_20250522"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20250825 object`¶

- `type: "code_execution_20250825"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20260120 object`¶

Code execution tool with REPL state persistence (daemon mode + gVisor checkpoint).¶

- `type: "code_execution_20260120"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20260521 object`¶

Code execution tool with REPL state persistence.¶

- `type: "code_execution_20260521"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `BrowserToolset20260801 object`¶

The browser toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the browser tool family. The model is served¶
the family's tool with any members disabled via `configs` removed¶
from its schema.¶

- `type: "browser_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `configs: optional BrowserToolsetConfigs or null`¶

Per-member configuration for `browser_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional BrowserTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `close_tab: optional BrowserCloseTabConfig or null`¶

`close_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional BrowserDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `file_upload: optional BrowserFileUploadConfig or null`¶

`file_upload`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `find: optional BrowserFindConfig or null`¶

`find`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `form_input: optional BrowserFormInputConfig or null`¶

`form_input`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `get_page_text: optional BrowserGetPageTextConfig or null`¶

`get_page_text`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional BrowserHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hover: optional BrowserHoverConfig or null`¶

`hover`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `javascript_exec: optional BrowserJavascriptExecConfig or null`¶

`javascript_exec`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional BrowserKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional BrowserLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional BrowserLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional BrowserLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional BrowserLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `list_tabs: optional BrowserListTabsConfig or null`¶

`list_tabs`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional BrowserMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional BrowserMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `navigate: optional BrowserNavigateConfig or null`¶

`navigate`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `new_tab: optional BrowserNewTabConfig or null`¶

`new_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_console: optional BrowserReadConsoleConfig or null`¶

`read_console`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_network: optional BrowserReadNetworkConfig or null`¶

`read_network`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_page: optional BrowserReadPageConfig or null`¶

`read_page`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional BrowserRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional BrowserScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional BrowserScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll_to: optional BrowserScrollToConfig or null`¶

`scroll_to`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `switch_tab: optional BrowserSwitchTabConfig or null`¶

`switch_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional BrowserTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional BrowserWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional BrowserZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `MemoryTool20250818 object`¶

- `type: "memory_20250818"`¶

- `name: "memory"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ComputerToolset20260801 object`¶

The computer toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the computer tool family. The model is¶
served the family's tool with any members disabled via `configs`¶
removed from its schema. Every member is enabled by default, zoom¶
included. The single-tool options `display_number` and¶
`enable_zoom` are not fields of a toolset entry — it carries only¶
`type`, `configs`, and `cache_control`; zoom is controlled¶
via `configs.zoom.enabled`.¶

- `type: "computer_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `configs: optional ComputerToolsetConfigs or null`¶

Per-member configuration for `computer_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional ComputerTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `cursor_position: optional ComputerCursorPositionConfig or null`¶

`cursor_position`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional ComputerDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional ComputerHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional ComputerKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional ComputerLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional ComputerLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional ComputerLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional ComputerLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional ComputerMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional ComputerMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional ComputerRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional ComputerScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional ComputerScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional ComputerTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional ComputerWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional ComputerZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `ToolTextEditor20250124 object`¶

- `type: "text_editor_20250124"`¶

- `name: "str_replace_editor"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolTextEditor20250429 object`¶

- `type: "text_editor_20250429"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolTextEditor20250728 object`¶

- `type: "text_editor_20250728"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `max_characters: optional number or null`¶

Maximum number of characters to display when viewing a file. If not specified, defaults to displaying the full file.¶

minimum: 1¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `WebSearchTool20250305 object`¶

- `type: "web_search_20250305"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `type: "approximate"`¶

- `city: optional string or null`¶

The city of the user.¶

maxLength: 255, minLength: 1¶

- `country: optional string or null`¶

The two letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2) of the user.¶

maxLength: 2, minLength: 2¶

- `region: optional string or null`¶

The region of the user.¶

maxLength: 255, minLength: 1¶

- `timezone: optional string or null`¶

The [IANA timezone](https://nodatime.org/TimeZones) of the user.¶

maxLength: 255, minLength: 1¶

- `WebFetchTool20250910 object`¶

- `type: "web_fetch_20250910"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `enabled: optional boolean`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

- `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`¶

Whether URLs in user messages are fetchable: "all" or "none".¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

 - `WebSearchTool20260209 object`¶

- `type: "web_search_20260209"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `WebFetchTool20260209 object`¶

- `type: "web_fetch_20260209"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `WebFetchTool20260309 object`¶

Web fetch tool with use_cache parameter for bypassing cached content.¶

- `type: "web_fetch_20260309"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

- `WebSearchTool20260318 object`¶

- `type: "web_search_20260318"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `WebFetchTool20260318 object`¶

- `type: "web_fetch_20260318"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

- `ToolSearchToolBm25_20251119 object`¶

- `type: "tool_search_tool_bm25_20251119" or "tool_search_tool_bm25"`¶

- `"tool_search_tool_bm25_20251119"`¶

- `"tool_search_tool_bm25"`¶

- `name: "tool_search_tool_bm25"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolSearchToolRegex20251119 object`¶

- `type: "tool_search_tool_regex_20251119" or "tool_search_tool_regex"`¶

- `"tool_search_tool_regex_20251119"`¶

- `"tool_search_tool_regex"`¶

- `name: "tool_search_tool_regex"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Message Create Params Container¶

- `MessageCreateParamsContainer = ContainerParams or string`¶

Container identifier for reuse across requests.¶

- `ContainerParams object`¶

Container parameters with skills to be loaded.¶

- `id: optional string or null`¶

Container id¶

- `skills: optional array of SkillParams or null`¶

List of skills to load in the container¶

maxItems: 20¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: optional string`¶

Skill version or 'latest' for most recent version¶

maxLength: 64, minLength: 1¶

- `string`¶

### Message Delta Usage¶

- `MessageDeltaUsage object`¶

- `cache_creation_input_tokens: number or null`¶

The cumulative number of input tokens used to create the cache entry.¶

minimum: 0¶

- `cache_read_input_tokens: number or null`¶

The cumulative number of input tokens read from the cache.¶

minimum: 0¶

- `input_tokens: number or null`¶

The cumulative number of input tokens which were used.¶

minimum: 0¶

- `output_tokens: number`¶

The cumulative number of output tokens which were used.¶

- `output_tokens_details: OutputTokensDetails or null`¶

Breakdown of output tokens by category.¶

`output_tokens` remains the inclusive, authoritative total used for billing.¶
This object provides a read-only decomposition for observability — for example,¶
how many of the billed output tokens were spent on internal reasoning that may¶
have been summarized before being returned to you.¶

- `thinking_tokens: number`¶

Number of output tokens the model generated as internal reasoning, including¶
the thinking-block delimiter tokens.¶

Reflects the raw reasoning the model produced, not the (possibly shorter)¶
summarized thinking text returned in the response body. Computed by¶
re-tokenizing the raw reasoning text, so it may differ from the model's exact¶
generation count by a small number of tokens. Always ≤ `output_tokens`;¶
`output_tokens - thinking_tokens` approximates the non-reasoning output.¶

default: 0, minimum: 0¶

- `server_tool_use: ServerToolUsage or null`¶

The number of server tool requests.¶

- `web_fetch_requests: number`¶

The number of web fetch tool requests.¶

default: 0, minimum: 0¶

- `web_search_requests: number`¶

The number of web search tool requests.¶

default: 0, minimum: 0¶

### Message Param¶

- `MessageParam object`¶

- `content: string or array of ContentBlockParam`¶

- `string`¶

- `array of ContentBlockParam`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

- `DocumentBlockParam object`¶

- `type: "document"`¶

- `source: Base64PDFSource or PlainTextSource or ContentBlockSource or 2 more`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `ContentBlockSource object`¶

- `type: "content"`¶

- `content: string or array of ContentBlockSourceContent`¶

- `string`¶

- `ContentBlockSourceContent = array of ContentBlockSourceContent`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `URLPDFSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileDocumentSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

- `enabled: optional boolean`¶

- `context: optional string or null`¶

minLength: 1¶

- `title: optional string or null`¶

maxLength: 500, minLength: 1¶

- `SearchResultBlockParam object`¶

- `type: "search_result"`¶

- `content: array of TextBlockParam`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional array of TextCitationParam or null`¶

- `source: string`¶

- `title: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam`¶

- `ThinkingBlockParam object`¶

- `type: "thinking"`¶

- `signature: string`¶

The `signature` value of this thinking block, exactly as returned by the API in a previous response. Used to verify that the block was generated by Claude.¶

Thinking blocks must be passed back unmodified and in their original order; a modified block results in a 400 `invalid_request_error`.¶

- `thinking: string`¶

The `thinking` text of this block as returned by the API.¶

- `RedactedThinkingBlockParam object`¶

- `type: "redacted_thinking"`¶

- `data: string`¶

The `data` value of this redacted thinking block, exactly as returned by the API in a previous response. Opaque and encrypted; pass it back unchanged.¶

- `ToolUseBlockParam object`¶

- `type: "tool_use"`¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `input: map[unknown]`¶

- `name: string`¶

maxLength: 200, minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family this member belongs to.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ToolResultBlockParam object`¶

- `type: "tool_result"`¶

- `tool_use_id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `content: optional string or array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `string`¶

- `array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `SearchResultBlockParam object`¶

- `DocumentBlockParam object`¶

- `ToolReferenceBlockParam object`¶

Tool reference block that can be included in tool_result content.¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BrowserStateBlockParam object`¶

The caller's browser state after a browser toolset member call —¶
the full inventory of open tabs, which tab is active, and any side¶
effects (tabs opened, download state changes) the call produced.¶

At most one per `tool_result`, only on a non-error result answering a¶
browser toolset member `tool_use`. The server renders the¶
model-visible text from it; the model never sees the raw fields.¶

- `type: "browser_state"`¶

- `tabs: array of BrowserStateTabEntry`¶

All tabs open in the browser after this call — the full inventory, not a delta. May be empty. Whenever non-empty, exactly one entry carries `active: true`.¶

maxItems: 100¶

- `tab_id: string`¶

The caller-assigned identifier for this tab, unique within the inventory.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `title: string`¶

The title of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The URL of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `active: optional boolean`¶

Whether this tab is the active tab after this call. Whenever `tabs` is non-empty, exactly one entry is marked `active: true`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `state_changes: optional array of BrowserStateChange or null`¶

Tabs opened and download state changes during this call. "Nothing to report" is expressed by omitting the field, never by an empty list.¶

maxItems: 200, minItems: 1¶

- `BrowserStateChangeTabOpened object`¶

A tab this call's execution opened that remains open at its end —¶
the creation delta of the `tabs` inventory, not an event log.¶

Carries only the `tab_id`; the tab's `title` and `url` live on its¶
`tabs` entry, which must include the same `tab_id`. A tab opened¶
during a failed call gets no deferred `tab_opened`; it simply appears¶
in the next result's `tabs` inventory.¶

- `type: "tab_opened"`¶

- `tab_id: string`¶

The `tab_id` of the opened tab, present in `tabs`.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadStarted object`¶

A file download that started during this call.¶

- `type: "download_started"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadCompleted object`¶

A file download that finished during this call, reported with the¶
same `download_id` as its `download_started` — or without a prior¶
`download_started`, when the download finished during the call that¶
started it (at most one state change per `download_id` per result).¶

- `type: "download_completed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `path: optional string or null`¶

Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `size_bytes: optional number or null`¶

The completed download's size.¶

minimum: 0¶

- `BrowserStateChangeDownloadFailed object`¶

A file download that failed — or was cancelled — during this call.¶

- `type: "download_failed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `error: optional string or null`¶

The failure or cancellation detail, when known.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `is_error: optional boolean`¶

- `toolset_name: optional string or null`¶

For a toolset member tool_result, the toolset family of the paired tool_use.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlockParam object`¶

- `type: "server_tool_use"`¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `WebSearchToolResultBlockParam object`¶

- `type: "web_search_tool_result"`¶

- `content: WebSearchToolResultBlockParamContent`¶

- `WebSearchToolResultBlockItem = array of WebSearchResultBlockParam`¶

- `type: "web_search_result"`¶

- `encrypted_content: string`¶

- `title: string`¶

- `url: string`¶

- `page_age: optional string or null`¶

- `WebSearchToolRequestError object`¶

- `type: "web_search_tool_result_error"`¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `WebFetchToolResultBlockParam object`¶

- `type: "web_fetch_tool_result"`¶

- `content: WebFetchToolResultErrorBlockParam or WebFetchBlockParam`¶

- `WebFetchToolResultErrorBlockParam object`¶

- `type: "web_fetch_tool_result_error"`¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlockParam object`¶

- `type: "web_fetch_result"`¶

- `content: DocumentBlockParam`¶

- `url: string`¶

Fetched content URL¶

- `retrieved_at: optional string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `CodeExecutionToolResultBlockParam object`¶

- `type: "code_execution_tool_result"`¶

- `content: CodeExecutionToolResultBlockParamContent`¶

- `CodeExecutionToolResultErrorParam object`¶

- `type: "code_execution_tool_result_error"`¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlockParam object`¶

- `type: "code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlockParam object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BashCodeExecutionToolResultBlockParam object`¶

- `type: "bash_code_execution_tool_result"`¶

- `content: BashCodeExecutionToolResultErrorParam or BashCodeExecutionResultBlockParam`¶

- `BashCodeExecutionToolResultErrorParam object`¶

- `type: "bash_code_execution_tool_result_error"`¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlockParam object`¶

- `type: "bash_code_execution_result"`¶

- `content: array of BashCodeExecutionOutputBlockParam`¶

- `type: "bash_code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `TextEditorCodeExecutionToolResultBlockParam object`¶

- `type: "text_editor_code_execution_tool_result"`¶

- `content: TextEditorCodeExecutionToolResultErrorParam or TextEditorCodeExecutionViewResultBlockParam or TextEditorCodeExecutionCreateResultBlockParam or TextEditorCodeExecutionStrReplaceResultBlockParam`¶

- `TextEditorCodeExecutionToolResultErrorParam object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: optional string or null`¶

- `TextEditorCodeExecutionViewResultBlockParam object`¶

- `type: "text_editor_code_execution_view_result"`¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: optional number or null`¶

- `start_line: optional number or null`¶

- `total_lines: optional number or null`¶

- `TextEditorCodeExecutionCreateResultBlockParam object`¶

- `type: "text_editor_code_execution_create_result"`¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlockParam object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

- `lines: optional array of string or null`¶

- `new_lines: optional number or null`¶

- `new_start: optional number or null`¶

- `old_lines: optional number or null`¶

- `old_start: optional number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `ToolSearchToolResultBlockParam object`¶

- `type: "tool_search_tool_result"`¶

- `content: ToolSearchToolResultErrorParam or ToolSearchToolSearchResultBlockParam`¶

- `ToolSearchToolResultErrorParam object`¶

- `type: "tool_search_tool_result_error"`¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: optional string or null`¶

- `ToolSearchToolSearchResultBlockParam object`¶

- `type: "tool_search_tool_search_result"`¶

- `tool_references: array of ToolReferenceBlockParam`¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `ContainerUploadBlockParam object`¶

A content block that represents a file to be uploaded to the container¶
Files uploaded via this block will be available in the container's input directory.¶

- `type: "container_upload"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `role: "user" or "assistant" or "system"`¶

- `"user"`¶

- `"assistant"`¶

- `"system"`¶

### Message Tokens Count¶

- `MessageTokensCount object`¶

- `input_tokens: number`¶

The total number of tokens across the provided list of messages, system prompt, and tools.¶

### Metadata¶

- `Metadata object`¶

- `user_id: optional string or null`¶

An external identifier for the user who is associated with the request.¶

This should be a uuid, hash value, or other opaque identifier. Anthropic may use this id to help detect abuse. Do not include any identifying information such as name, email address, or phone number.¶

maxLength: 512¶

### Model¶

- `Model = "claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more or string`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1"`¶

Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows¶

- `"claude-mythos-5-1"`¶

Our most capable model for cybersecurity and biology research, available through trusted access programs¶

- `"claude-sonnet-5"`¶

High-performance model for coding and agents¶

- `"claude-fable-5"`¶

Next generation of intelligence for the hardest knowledge work and coding problems¶

- `"claude-mythos-5"`¶

Most capable model for cybersecurity and biology research¶

- `"claude-opus-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-8"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-7"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-mythos-preview"`¶

New class of intelligence, strongest in coding and cybersecurity¶

- `"claude-opus-4-6"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-6"`¶

Best combination of speed and intelligence¶

- `"claude-haiku-4-5"`¶

Fastest model with near-frontier intelligence¶

- `"claude-haiku-4-5-20251001"`¶

Fastest model with near-frontier intelligence¶

- `"claude-opus-4-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-5-20251101"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-5"`¶

High-performance model for agents and coding¶

- `"claude-sonnet-4-5-20250929"`¶

High-performance model for agents and coding¶

- `string`¶

### Output Config¶

- `OutputConfig object`¶

- `effort: optional "low" or "medium" or "high" or 2 more or null`¶

All possible effort levels.¶

- `"low"`¶

- `"medium"`¶

- `"high"`¶

- `"xhigh"`¶

- `"max"`¶

- `format: optional JSONOutputFormat or null`¶

A schema to specify Claude's output format in responses. See [structured outputs](https://platform.claude.com/docs/en/build-with-claude/structured-outputs)¶

- `type: "json_schema"`¶

- `schema: map[unknown]`¶

The JSON schema of the format¶

### Output Tokens Details¶

- `OutputTokensDetails object`¶

- `thinking_tokens: number`¶

Number of output tokens the model generated as internal reasoning, including¶
the thinking-block delimiter tokens.¶

Reflects the raw reasoning the model produced, not the (possibly shorter)¶
summarized thinking text returned in the response body. Computed by¶
re-tokenizing the raw reasoning text, so it may differ from the model's exact¶
generation count by a small number of tokens. Always ≤ `output_tokens`;¶
`output_tokens - thinking_tokens` approximates the non-reasoning output.¶

default: 0, minimum: 0¶

### Plain Text Source¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

### Raw Content Block Delta¶

- `RawContentBlockDelta = TextDelta or InputJSONDelta or CitationsDelta or 2 more`¶

- `TextDelta object`¶

- `type: "text_delta"`¶

default: text_delta¶

- `text: string`¶

- `InputJSONDelta object`¶

- `type: "input_json_delta"`¶

default: input_json_delta¶

- `partial_json: string`¶

- `CitationsDelta object`¶

- `type: "citations_delta"`¶

default: citations_delta¶

- `citation: CitationCharLocation or CitationPageLocation or CitationContentBlockLocation or 2 more`¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ThinkingDelta object`¶

- `type: "thinking_delta"`¶

default: thinking_delta¶

- `thinking: string`¶

The incremental `thinking` text for this content block. Concatenate the `thinking` values of successive `thinking_delta` events to assemble the block's full `thinking` value.¶

- `SignatureDelta object`¶

- `type: "signature_delta"`¶

default: signature_delta¶

- `signature: string`¶

The `signature` for this thinking block: an opaque value used to verify that the block was generated by Claude when it is passed back to the API. Delivered in a `signature_delta` event just before the block's `content_block_stop` event.¶

### Raw Content Block Delta Event¶

- `RawContentBlockDeltaEvent object`¶

- `type: "content_block_delta"`¶

default: content_block_delta¶

- `delta: RawContentBlockDelta`¶

- `TextDelta object`¶

- `type: "text_delta"`¶

default: text_delta¶

- `text: string`¶

- `InputJSONDelta object`¶

- `type: "input_json_delta"`¶

default: input_json_delta¶

- `partial_json: string`¶

- `CitationsDelta object`¶

- `type: "citations_delta"`¶

default: citations_delta¶

- `citation: CitationCharLocation or CitationPageLocation or CitationContentBlockLocation or 2 more`¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ThinkingDelta object`¶

- `type: "thinking_delta"`¶

default: thinking_delta¶

- `thinking: string`¶

The incremental `thinking` text for this content block. Concatenate the `thinking` values of successive `thinking_delta` events to assemble the block's full `thinking` value.¶

- `SignatureDelta object`¶

- `type: "signature_delta"`¶

default: signature_delta¶

- `signature: string`¶

The `signature` for this thinking block: an opaque value used to verify that the block was generated by Claude when it is passed back to the API. Delivered in a `signature_delta` event just before the block's `content_block_stop` event.¶

- `index: number`¶

### Raw Content Block Start Event¶

- `RawContentBlockStartEvent object`¶

- `type: "content_block_start"`¶

default: content_block_start¶

- `content_block: TextBlock or ThinkingBlock or RedactedThinkingBlock or 9 more`¶

- `TextBlock object`¶

- `type: "text"`¶

default: text¶

- `citations: array of TextCitation or null`¶

Citations supporting the text block.¶

The type of citation returned will depend on the type of document being cited. Citing a PDF results in `page_location`, plain text results in `char_location`, and content document results in `content_block_location`.¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `text: string`¶

minLength: 0¶

- `ThinkingBlock object`¶

- `type: "thinking"`¶

default: thinking¶

- `signature: string`¶

A value used to verify that this thinking block was generated by Claude when it is passed back to the API.¶

This is an opaque field and should not be interpreted or parsed. When passing thinking blocks back to the API (required when using tools with extended thinking), pass them back exactly as received, with this field intact.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `thinking: string`¶

The text of Claude's thinking process for this block.¶

- `RedactedThinkingBlock object`¶

- `type: "redacted_thinking"`¶

default: redacted_thinking¶

- `data: string`¶

The contents of this redacted thinking block, returned when portions of the model's thinking were safety-redacted. This field is opaque and encrypted, with no readable content.¶

Pass `redacted_thinking` blocks back to the API unchanged when continuing a multi-turn conversation.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking#redacted-thinking-blocks) for details.¶

- `ToolUseBlock object`¶

- `type: "tool_use"`¶

default: tool_use¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: string`¶

minLength: 1¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlock object`¶

- `type: "server_tool_use"`¶

default: server_tool_use¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `WebSearchToolResultBlock object`¶

- `type: "web_search_tool_result"`¶

default: web_search_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebSearchToolResultBlockContent`¶

- `WebSearchToolResultError object`¶

- `type: "web_search_tool_result_error"`¶

default: web_search_tool_result_error¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `array of WebSearchResultBlock`¶

- `type: "web_search_result"`¶

default: web_search_result¶

- `encrypted_content: string`¶

- `page_age: string or null`¶

- `title: string`¶

- `url: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `WebFetchToolResultBlock object`¶

- `type: "web_fetch_tool_result"`¶

default: web_fetch_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebFetchToolResultErrorBlock or WebFetchBlock`¶

- `WebFetchToolResultErrorBlock object`¶

- `type: "web_fetch_tool_result_error"`¶

default: web_fetch_tool_result_error¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlock object`¶

- `type: "web_fetch_result"`¶

default: web_fetch_result¶

- `content: DocumentBlock`¶

- `type: "document"`¶

default: document¶

- `citations: CitationsConfig or null`¶

Citation configuration for the document¶

- `enabled: boolean`¶

default: false¶

- `source: Base64PDFSource or PlainTextSource`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `title: string or null`¶

The title of the document¶

- `retrieved_at: string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `url: string`¶

Fetched content URL¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `CodeExecutionToolResultBlock object`¶

- `type: "code_execution_tool_result"`¶

default: code_execution_tool_result¶

- `content: CodeExecutionToolResultBlockContent`¶

- `CodeExecutionToolResultError object`¶

- `type: "code_execution_tool_result_error"`¶

default: code_execution_tool_result_error¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlock object`¶

- `type: "code_execution_result"`¶

default: code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlock object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

default: encrypted_code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `BashCodeExecutionToolResultBlock object`¶

- `type: "bash_code_execution_tool_result"`¶

default: bash_code_execution_tool_result¶

- `content: BashCodeExecutionToolResultError or BashCodeExecutionResultBlock`¶

- `BashCodeExecutionToolResultError object`¶

- `type: "bash_code_execution_tool_result_error"`¶

default: bash_code_execution_tool_result_error¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlock object`¶

- `type: "bash_code_execution_result"`¶

default: bash_code_execution_result¶

- `content: array of BashCodeExecutionOutputBlock`¶

- `type: "bash_code_execution_output"`¶

default: bash_code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `TextEditorCodeExecutionToolResultBlock object`¶

- `type: "text_editor_code_execution_tool_result"`¶

default: text_editor_code_execution_tool_result¶

- `content: TextEditorCodeExecutionToolResultError or TextEditorCodeExecutionViewResultBlock or TextEditorCodeExecutionCreateResultBlock or TextEditorCodeExecutionStrReplaceResultBlock`¶

- `TextEditorCodeExecutionToolResultError object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

default: text_editor_code_execution_tool_result_error¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: string or null`¶

- `TextEditorCodeExecutionViewResultBlock object`¶

- `type: "text_editor_code_execution_view_result"`¶

default: text_editor_code_execution_view_result¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: number or null`¶

- `start_line: number or null`¶

- `total_lines: number or null`¶

- `TextEditorCodeExecutionCreateResultBlock object`¶

- `type: "text_editor_code_execution_create_result"`¶

default: text_editor_code_execution_create_result¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlock object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

default: text_editor_code_execution_str_replace_result¶

- `lines: array of string or null`¶

- `new_lines: number or null`¶

- `new_start: number or null`¶

- `old_lines: number or null`¶

- `old_start: number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ToolSearchToolResultBlock object`¶

- `type: "tool_search_tool_result"`¶

default: tool_search_tool_result¶

- `content: ToolSearchToolResultError or ToolSearchToolSearchResultBlock`¶

- `ToolSearchToolResultError object`¶

- `type: "tool_search_tool_result_error"`¶

default: tool_search_tool_result_error¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: string or null`¶

- `ToolSearchToolSearchResultBlock object`¶

- `type: "tool_search_tool_search_result"`¶

default: tool_search_tool_search_result¶

- `tool_references: array of ToolReferenceBlock`¶

- `type: "tool_reference"`¶

default: tool_reference¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ContainerUploadBlock object`¶

Response model for a file uploaded to the container.¶

- `type: "container_upload"`¶

default: container_upload¶

- `file_id: string`¶

- `index: number`¶

### Raw Content Block Stop Event¶

- `RawContentBlockStopEvent object`¶

- `type: "content_block_stop"`¶

default: content_block_stop¶

- `index: number`¶

### Raw Message Delta Event¶

- `RawMessageDeltaEvent object`¶

- `type: "message_delta"`¶

default: message_delta¶

- `delta: object`¶

- `container: Container or null`¶

Information about the container used in the request (for the code execution tool)¶

- `id: string`¶

Identifier for the container used in this request¶

- `expires_at: string`¶

The time at which the container will expire.¶

format: date-time¶

- `skills: array of ContainerSkill or null`¶

Skills loaded in the container¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: string`¶

The resolved version: a skill version ID for custom skills.¶

maxLength: 64, minLength: 1¶

- `stop_details: RefusalStopDetails or null`¶

Structured information about a refusal.¶

- `type: "refusal"`¶

default: refusal¶

- `category: "cyber" or "bio" or "frontier_llm" or 2 more or null`¶

The policy category that triggered a refusal.¶

- `"cyber"`¶

The request could enable cyber harm, such as malware or exploit development. Benign cybersecurity work can also trigger this category.¶

- `"bio"`¶

The request could enable biological harm, such as dangerous lab methods. Beneficial life sciences work can also trigger this category.¶

- `"frontier_llm"`¶

The request could assist the development of competing AI models, which is restricted under [Anthropic's commercial terms](https://www.anthropic.com/legal/commercial-terms). Benign machine learning work can also trigger this category.¶

- `"reasoning_extraction"`¶

The request asks the model to reproduce its internal reasoning in the response text. To get reasoning in a structured form instead, use [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/adaptive-thinking).¶

- `"general_harms"`¶

The request could be related to an area that was determined as harmful. Benign work might sometimes trigger this category.¶

- `explanation: string or null`¶

Human-readable explanation of the refusal.¶

This text is not guaranteed to be stable. `null` when no explanation is available for the category.¶

- `stop_reason: StopReason or null`¶

- `"end_turn"`¶

- `"max_tokens"`¶

- `"stop_sequence"`¶

- `"tool_use"`¶

- `"pause_turn"`¶

- `"refusal"`¶

- `"model_context_window_exceeded"`¶

- `stop_sequence: string or null`¶

- `usage: MessageDeltaUsage`¶

Billing and rate-limit usage.¶

Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.¶

Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.¶

For example, `output_tokens` will be non-zero, even for an empty string response from Claude.¶

Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.¶

- `cache_creation_input_tokens: number or null`¶

The cumulative number of input tokens used to create the cache entry.¶

minimum: 0¶

- `cache_read_input_tokens: number or null`¶

The cumulative number of input tokens read from the cache.¶

minimum: 0¶

- `input_tokens: number or null`¶

The cumulative number of input tokens which were used.¶

minimum: 0¶

- `output_tokens: number`¶

The cumulative number of output tokens which were used.¶

- `output_tokens_details: OutputTokensDetails or null`¶

Breakdown of output tokens by category.¶

`output_tokens` remains the inclusive, authoritative total used for billing.¶
This object provides a read-only decomposition for observability — for example,¶
how many of the billed output tokens were spent on internal reasoning that may¶
have been summarized before being returned to you.¶

- `thinking_tokens: number`¶

Number of output tokens the model generated as internal reasoning, including¶
the thinking-block delimiter tokens.¶

Reflects the raw reasoning the model produced, not the (possibly shorter)¶
summarized thinking text returned in the response body. Computed by¶
re-tokenizing the raw reasoning text, so it may differ from the model's exact¶
generation count by a small number of tokens. Always ≤ `output_tokens`;¶
`output_tokens - thinking_tokens` approximates the non-reasoning output.¶

default: 0, minimum: 0¶

- `server_tool_use: ServerToolUsage or null`¶

The number of server tool requests.¶

- `web_fetch_requests: number`¶

The number of web fetch tool requests.¶

default: 0, minimum: 0¶

- `web_search_requests: number`¶

The number of web search tool requests.¶

default: 0, minimum: 0¶

### Raw Message Start Event¶

- `RawMessageStartEvent object`¶

- `type: "message_start"`¶

default: message_start¶

- `message: Message`¶

- `type: "message"`¶

Object type.¶

For Messages, this is always `"message"`.¶

default: message¶

- `id: string`¶

Unique object identifier.¶

The format and length of IDs may change over time.¶

- `container: Container or null`¶

Information about the container used in the request (for the code execution tool)¶

- `id: string`¶

Identifier for the container used in this request¶

- `expires_at: string`¶

The time at which the container will expire.¶

format: date-time¶

- `skills: array of ContainerSkill or null`¶

Skills loaded in the container¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: string`¶

The resolved version: a skill version ID for custom skills.¶

maxLength: 64, minLength: 1¶

- `content: array of ContentBlock`¶

Content generated by the model.¶

This is an array of content blocks, each of which has a `type` that determines its shape.¶

Example:¶

```json¶
[{"type": "text", "text": "Hi, I'm Claude."}]¶
```¶

If the request input `messages` ended with an `assistant` turn, then the response `content` will continue directly from that last turn. You can use this to constrain the model's output.¶

For example, if the input `messages` were:¶

```json¶
[¶
{"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},¶
{"role": "assistant", "content": "The best answer is ("}¶
]¶
```¶

Then the response `content` might be:¶

```json¶
[{"type": "text", "text": "B)"}]¶
```¶

- `TextBlock object`¶

- `type: "text"`¶

default: text¶

- `citations: array of TextCitation or null`¶

Citations supporting the text block.¶

The type of citation returned will depend on the type of document being cited. Citing a PDF results in `page_location`, plain text results in `char_location`, and content document results in `content_block_location`.¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `text: string`¶

minLength: 0¶

- `ThinkingBlock object`¶

- `type: "thinking"`¶

default: thinking¶

- `signature: string`¶

A value used to verify that this thinking block was generated by Claude when it is passed back to the API.¶

This is an opaque field and should not be interpreted or parsed. When passing thinking blocks back to the API (required when using tools with extended thinking), pass them back exactly as received, with this field intact.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `thinking: string`¶

The text of Claude's thinking process for this block.¶

- `RedactedThinkingBlock object`¶

- `type: "redacted_thinking"`¶

default: redacted_thinking¶

- `data: string`¶

The contents of this redacted thinking block, returned when portions of the model's thinking were safety-redacted. This field is opaque and encrypted, with no readable content.¶

Pass `redacted_thinking` blocks back to the API unchanged when continuing a multi-turn conversation.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking#redacted-thinking-blocks) for details.¶

- `ToolUseBlock object`¶

- `type: "tool_use"`¶

default: tool_use¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: string`¶

minLength: 1¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlock object`¶

- `type: "server_tool_use"`¶

default: server_tool_use¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `WebSearchToolResultBlock object`¶

- `type: "web_search_tool_result"`¶

default: web_search_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebSearchToolResultBlockContent`¶

- `WebSearchToolResultError object`¶

- `type: "web_search_tool_result_error"`¶

default: web_search_tool_result_error¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `array of WebSearchResultBlock`¶

- `type: "web_search_result"`¶

default: web_search_result¶

- `encrypted_content: string`¶

- `page_age: string or null`¶

- `title: string`¶

- `url: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `WebFetchToolResultBlock object`¶

- `type: "web_fetch_tool_result"`¶

default: web_fetch_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebFetchToolResultErrorBlock or WebFetchBlock`¶

- `WebFetchToolResultErrorBlock object`¶

- `type: "web_fetch_tool_result_error"`¶

default: web_fetch_tool_result_error¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlock object`¶

- `type: "web_fetch_result"`¶

default: web_fetch_result¶

- `content: DocumentBlock`¶

- `type: "document"`¶

default: document¶

- `citations: CitationsConfig or null`¶

Citation configuration for the document¶

- `enabled: boolean`¶

default: false¶

- `source: Base64PDFSource or PlainTextSource`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `title: string or null`¶

The title of the document¶

- `retrieved_at: string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `url: string`¶

Fetched content URL¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `CodeExecutionToolResultBlock object`¶

- `type: "code_execution_tool_result"`¶

default: code_execution_tool_result¶

- `content: CodeExecutionToolResultBlockContent`¶

- `CodeExecutionToolResultError object`¶

- `type: "code_execution_tool_result_error"`¶

default: code_execution_tool_result_error¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlock object`¶

- `type: "code_execution_result"`¶

default: code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlock object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

default: encrypted_code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `BashCodeExecutionToolResultBlock object`¶

- `type: "bash_code_execution_tool_result"`¶

default: bash_code_execution_tool_result¶

- `content: BashCodeExecutionToolResultError or BashCodeExecutionResultBlock`¶

- `BashCodeExecutionToolResultError object`¶

- `type: "bash_code_execution_tool_result_error"`¶

default: bash_code_execution_tool_result_error¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlock object`¶

- `type: "bash_code_execution_result"`¶

default: bash_code_execution_result¶

- `content: array of BashCodeExecutionOutputBlock`¶

- `type: "bash_code_execution_output"`¶

default: bash_code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `TextEditorCodeExecutionToolResultBlock object`¶

- `type: "text_editor_code_execution_tool_result"`¶

default: text_editor_code_execution_tool_result¶

- `content: TextEditorCodeExecutionToolResultError or TextEditorCodeExecutionViewResultBlock or TextEditorCodeExecutionCreateResultBlock or TextEditorCodeExecutionStrReplaceResultBlock`¶

- `TextEditorCodeExecutionToolResultError object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

default: text_editor_code_execution_tool_result_error¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: string or null`¶

- `TextEditorCodeExecutionViewResultBlock object`¶

- `type: "text_editor_code_execution_view_result"`¶

default: text_editor_code_execution_view_result¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: number or null`¶

- `start_line: number or null`¶

- `total_lines: number or null`¶

- `TextEditorCodeExecutionCreateResultBlock object`¶

- `type: "text_editor_code_execution_create_result"`¶

default: text_editor_code_execution_create_result¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlock object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

default: text_editor_code_execution_str_replace_result¶

- `lines: array of string or null`¶

- `new_lines: number or null`¶

- `new_start: number or null`¶

- `old_lines: number or null`¶

- `old_start: number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ToolSearchToolResultBlock object`¶

- `type: "tool_search_tool_result"`¶

default: tool_search_tool_result¶

- `content: ToolSearchToolResultError or ToolSearchToolSearchResultBlock`¶

- `ToolSearchToolResultError object`¶

- `type: "tool_search_tool_result_error"`¶

default: tool_search_tool_result_error¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: string or null`¶

- `ToolSearchToolSearchResultBlock object`¶

- `type: "tool_search_tool_search_result"`¶

default: tool_search_tool_search_result¶

- `tool_references: array of ToolReferenceBlock`¶

- `type: "tool_reference"`¶

default: tool_reference¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ContainerUploadBlock object`¶

Response model for a file uploaded to the container.¶

- `type: "container_upload"`¶

default: container_upload¶

- `file_id: string`¶

- `model: Model`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1"`¶

Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows¶

- `"claude-mythos-5-1"`¶

Our most capable model for cybersecurity and biology research, available through trusted access programs¶

- `"claude-sonnet-5"`¶

High-performance model for coding and agents¶

- `"claude-fable-5"`¶

Next generation of intelligence for the hardest knowledge work and coding problems¶

- `"claude-mythos-5"`¶

Most capable model for cybersecurity and biology research¶

- `"claude-opus-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-8"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-7"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-mythos-preview"`¶

New class of intelligence, strongest in coding and cybersecurity¶

- `"claude-opus-4-6"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-6"`¶

Best combination of speed and intelligence¶

- `"claude-haiku-4-5"`¶

Fastest model with near-frontier intelligence¶

- `"claude-haiku-4-5-20251001"`¶

Fastest model with near-frontier intelligence¶

- `"claude-opus-4-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-5-20251101"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-5"`¶

High-performance model for agents and coding¶

- `"claude-sonnet-4-5-20250929"`¶

High-performance model for agents and coding¶

- `string`¶

- `role: "assistant"`¶

Conversational role of the generated message.¶

This will always be `"assistant"`.¶

default: assistant¶

- `stop_details: RefusalStopDetails or null`¶

Structured information about a refusal.¶

- `type: "refusal"`¶

default: refusal¶

- `category: "cyber" or "bio" or "frontier_llm" or 2 more or null`¶

The policy category that triggered a refusal.¶

- `"cyber"`¶

The request could enable cyber harm, such as malware or exploit development. Benign cybersecurity work can also trigger this category.¶

- `"bio"`¶

The request could enable biological harm, such as dangerous lab methods. Beneficial life sciences work can also trigger this category.¶

- `"frontier_llm"`¶

The request could assist the development of competing AI models, which is restricted under [Anthropic's commercial terms](https://www.anthropic.com/legal/commercial-terms). Benign machine learning work can also trigger this category.¶

- `"reasoning_extraction"`¶

The request asks the model to reproduce its internal reasoning in the response text. To get reasoning in a structured form instead, use [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/adaptive-thinking).¶

- `"general_harms"`¶

The request could be related to an area that was determined as harmful. Benign work might sometimes trigger this category.¶

- `explanation: string or null`¶

Human-readable explanation of the refusal.¶

This text is not guaranteed to be stable. `null` when no explanation is available for the category.¶

- `stop_reason: StopReason or null`¶

The reason that we stopped.¶

This may be one the following values:¶

* `"end_turn"`: the model reached a natural stopping point¶
* `"max_tokens"`: we exceeded the requested `max_tokens` or the model's maximum¶
* `"stop_sequence"`: one of your provided custom `stop_sequences` was generated¶
* `"tool_use"`: the model invoked one or more tools¶
* `"pause_turn"`: we paused a long-running turn. You may provide the response back as-is in a subsequent request to let the model continue.¶
* `"refusal"`: when streaming classifiers intervene to handle potential policy violations¶
* `"model_context_window_exceeded"`: we exceeded the model's context window¶

In non-streaming mode this value is always non-null. In streaming mode, it is null in the `message_start` event and non-null otherwise.¶

- `"end_turn"`¶

- `"max_tokens"`¶

- `"stop_sequence"`¶

- `"tool_use"`¶

- `"pause_turn"`¶

- `"refusal"`¶

- `"model_context_window_exceeded"`¶

- `stop_sequence: string or null`¶

Which custom stop sequence was generated, if any.¶

This value will be a non-null string if one of your custom stop sequences was generated.¶

- `usage: Usage`¶

Billing and rate-limit usage.¶

Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.¶

Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.¶

For example, `output_tokens` will be non-zero, even for an empty string response from Claude.¶

Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.¶

- `cache_creation: CacheCreation or null`¶

Breakdown of cached tokens by TTL¶

- `ephemeral_1h_input_tokens: number`¶

The number of input tokens used to create the 1 hour cache entry.¶

default: 0, minimum: 0¶

- `ephemeral_5m_input_tokens: number`¶

The number of input tokens used to create the 5 minute cache entry.¶

default: 0, minimum: 0¶

- `cache_creation_input_tokens: number or null`¶

The number of input tokens used to create the cache entry.¶

minimum: 0¶

- `cache_read_input_tokens: number or null`¶

The number of input tokens read from the cache.¶

minimum: 0¶

- `inference_geo: string or null`¶

The geographic region where inference was performed for this request.¶

- `input_tokens: number`¶

The number of input tokens which were used.¶

minimum: 0¶

- `output_tokens: number`¶

The number of output tokens which were used.¶

minimum: 0¶

- `output_tokens_details: OutputTokensDetails or null`¶

Breakdown of output tokens by category.¶

`output_tokens` remains the inclusive, authoritative total used for billing.¶
This object provides a read-only decomposition for observability — for example,¶
how many of the billed output tokens were spent on internal reasoning that may¶
have been summarized before being returned to you.¶

- `thinking_tokens: number`¶

Number of output tokens the model generated as internal reasoning, including¶
the thinking-block delimiter tokens.¶

Reflects the raw reasoning the model produced, not the (possibly shorter)¶
summarized thinking text returned in the response body. Computed by¶
re-tokenizing the raw reasoning text, so it may differ from the model's exact¶
generation count by a small number of tokens. Always ≤ `output_tokens`;¶
`output_tokens - thinking_tokens` approximates the non-reasoning output.¶

default: 0, minimum: 0¶

- `server_tool_use: ServerToolUsage or null`¶

The number of server tool requests.¶

- `web_fetch_requests: number`¶

The number of web fetch tool requests.¶

default: 0, minimum: 0¶

- `web_search_requests: number`¶

The number of web search tool requests.¶

default: 0, minimum: 0¶

- `service_tier: "standard" or "priority" or "batch" or null`¶

If the request used the priority, standard, or batch tier.¶

- `"standard"`¶

- `"priority"`¶

- `"batch"`¶

### Raw Message Stop Event¶

- `RawMessageStopEvent object`¶

- `type: "message_stop"`¶

default: message_stop¶

### Raw Message Stream Event¶

- `RawMessageStreamEvent = RawMessageStartEvent or RawMessageDeltaEvent or RawMessageStopEvent or 3 more`¶

- `RawMessageStartEvent object`¶

- `type: "message_start"`¶

default: message_start¶

- `message: Message`¶

- `type: "message"`¶

Object type.¶

For Messages, this is always `"message"`.¶

default: message¶

- `id: string`¶

Unique object identifier.¶

The format and length of IDs may change over time.¶

- `container: Container or null`¶

Information about the container used in the request (for the code execution tool)¶

- `id: string`¶

Identifier for the container used in this request¶

- `expires_at: string`¶

The time at which the container will expire.¶

format: date-time¶

- `skills: array of ContainerSkill or null`¶

Skills loaded in the container¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: string`¶

The resolved version: a skill version ID for custom skills.¶

maxLength: 64, minLength: 1¶

- `content: array of ContentBlock`¶

Content generated by the model.¶

This is an array of content blocks, each of which has a `type` that determines its shape.¶

Example:¶

```json¶
[{"type": "text", "text": "Hi, I'm Claude."}]¶
```¶

If the request input `messages` ended with an `assistant` turn, then the response `content` will continue directly from that last turn. You can use this to constrain the model's output.¶

For example, if the input `messages` were:¶

```json¶
[¶
{"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},¶
{"role": "assistant", "content": "The best answer is ("}¶
]¶
```¶

Then the response `content` might be:¶

```json¶
[{"type": "text", "text": "B)"}]¶
```¶

- `TextBlock object`¶

- `type: "text"`¶

default: text¶

- `citations: array of TextCitation or null`¶

Citations supporting the text block.¶

The type of citation returned will depend on the type of document being cited. Citing a PDF results in `page_location`, plain text results in `char_location`, and content document results in `content_block_location`.¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `text: string`¶

minLength: 0¶

- `ThinkingBlock object`¶

- `type: "thinking"`¶

default: thinking¶

- `signature: string`¶

A value used to verify that this thinking block was generated by Claude when it is passed back to the API.¶

This is an opaque field and should not be interpreted or parsed. When passing thinking blocks back to the API (required when using tools with extended thinking), pass them back exactly as received, with this field intact.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `thinking: string`¶

The text of Claude's thinking process for this block.¶

- `RedactedThinkingBlock object`¶

- `type: "redacted_thinking"`¶

default: redacted_thinking¶

- `data: string`¶

The contents of this redacted thinking block, returned when portions of the model's thinking were safety-redacted. This field is opaque and encrypted, with no readable content.¶

Pass `redacted_thinking` blocks back to the API unchanged when continuing a multi-turn conversation.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking#redacted-thinking-blocks) for details.¶

- `ToolUseBlock object`¶

- `type: "tool_use"`¶

default: tool_use¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: string`¶

minLength: 1¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlock object`¶

- `type: "server_tool_use"`¶

default: server_tool_use¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `WebSearchToolResultBlock object`¶

- `type: "web_search_tool_result"`¶

default: web_search_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebSearchToolResultBlockContent`¶

- `WebSearchToolResultError object`¶

- `type: "web_search_tool_result_error"`¶

default: web_search_tool_result_error¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `array of WebSearchResultBlock`¶

- `type: "web_search_result"`¶

default: web_search_result¶

- `encrypted_content: string`¶

- `page_age: string or null`¶

- `title: string`¶

- `url: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `WebFetchToolResultBlock object`¶

- `type: "web_fetch_tool_result"`¶

default: web_fetch_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebFetchToolResultErrorBlock or WebFetchBlock`¶

- `WebFetchToolResultErrorBlock object`¶

- `type: "web_fetch_tool_result_error"`¶

default: web_fetch_tool_result_error¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlock object`¶

- `type: "web_fetch_result"`¶

default: web_fetch_result¶

- `content: DocumentBlock`¶

- `type: "document"`¶

default: document¶

- `citations: CitationsConfig or null`¶

Citation configuration for the document¶

- `enabled: boolean`¶

default: false¶

- `source: Base64PDFSource or PlainTextSource`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `title: string or null`¶

The title of the document¶

- `retrieved_at: string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `url: string`¶

Fetched content URL¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `CodeExecutionToolResultBlock object`¶

- `type: "code_execution_tool_result"`¶

default: code_execution_tool_result¶

- `content: CodeExecutionToolResultBlockContent`¶

- `CodeExecutionToolResultError object`¶

- `type: "code_execution_tool_result_error"`¶

default: code_execution_tool_result_error¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlock object`¶

- `type: "code_execution_result"`¶

default: code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlock object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

default: encrypted_code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `BashCodeExecutionToolResultBlock object`¶

- `type: "bash_code_execution_tool_result"`¶

default: bash_code_execution_tool_result¶

- `content: BashCodeExecutionToolResultError or BashCodeExecutionResultBlock`¶

- `BashCodeExecutionToolResultError object`¶

- `type: "bash_code_execution_tool_result_error"`¶

default: bash_code_execution_tool_result_error¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlock object`¶

- `type: "bash_code_execution_result"`¶

default: bash_code_execution_result¶

- `content: array of BashCodeExecutionOutputBlock`¶

- `type: "bash_code_execution_output"`¶

default: bash_code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `TextEditorCodeExecutionToolResultBlock object`¶

- `type: "text_editor_code_execution_tool_result"`¶

default: text_editor_code_execution_tool_result¶

- `content: TextEditorCodeExecutionToolResultError or TextEditorCodeExecutionViewResultBlock or TextEditorCodeExecutionCreateResultBlock or TextEditorCodeExecutionStrReplaceResultBlock`¶

- `TextEditorCodeExecutionToolResultError object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

default: text_editor_code_execution_tool_result_error¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: string or null`¶

- `TextEditorCodeExecutionViewResultBlock object`¶

- `type: "text_editor_code_execution_view_result"`¶

default: text_editor_code_execution_view_result¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: number or null`¶

- `start_line: number or null`¶

- `total_lines: number or null`¶

- `TextEditorCodeExecutionCreateResultBlock object`¶

- `type: "text_editor_code_execution_create_result"`¶

default: text_editor_code_execution_create_result¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlock object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

default: text_editor_code_execution_str_replace_result¶

- `lines: array of string or null`¶

- `new_lines: number or null`¶

- `new_start: number or null`¶

- `old_lines: number or null`¶

- `old_start: number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ToolSearchToolResultBlock object`¶

- `type: "tool_search_tool_result"`¶

default: tool_search_tool_result¶

- `content: ToolSearchToolResultError or ToolSearchToolSearchResultBlock`¶

- `ToolSearchToolResultError object`¶

- `type: "tool_search_tool_result_error"`¶

default: tool_search_tool_result_error¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: string or null`¶

- `ToolSearchToolSearchResultBlock object`¶

- `type: "tool_search_tool_search_result"`¶

default: tool_search_tool_search_result¶

- `tool_references: array of ToolReferenceBlock`¶

- `type: "tool_reference"`¶

default: tool_reference¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ContainerUploadBlock object`¶

Response model for a file uploaded to the container.¶

- `type: "container_upload"`¶

default: container_upload¶

- `file_id: string`¶

- `model: Model`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1"`¶

Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows¶

- `"claude-mythos-5-1"`¶

Our most capable model for cybersecurity and biology research, available through trusted access programs¶

- `"claude-sonnet-5"`¶

High-performance model for coding and agents¶

- `"claude-fable-5"`¶

Next generation of intelligence for the hardest knowledge work and coding problems¶

- `"claude-mythos-5"`¶

Most capable model for cybersecurity and biology research¶

- `"claude-opus-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-8"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-7"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-mythos-preview"`¶

New class of intelligence, strongest in coding and cybersecurity¶

- `"claude-opus-4-6"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-6"`¶

Best combination of speed and intelligence¶

- `"claude-haiku-4-5"`¶

Fastest model with near-frontier intelligence¶

- `"claude-haiku-4-5-20251001"`¶

Fastest model with near-frontier intelligence¶

- `"claude-opus-4-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-5-20251101"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-5"`¶

High-performance model for agents and coding¶

- `"claude-sonnet-4-5-20250929"`¶

High-performance model for agents and coding¶

- `string`¶

- `role: "assistant"`¶

Conversational role of the generated message.¶

This will always be `"assistant"`.¶

default: assistant¶

- `stop_details: RefusalStopDetails or null`¶

Structured information about a refusal.¶

- `type: "refusal"`¶

default: refusal¶

- `category: "cyber" or "bio" or "frontier_llm" or 2 more or null`¶

The policy category that triggered a refusal.¶

- `"cyber"`¶

The request could enable cyber harm, such as malware or exploit development. Benign cybersecurity work can also trigger this category.¶

- `"bio"`¶

The request could enable biological harm, such as dangerous lab methods. Beneficial life sciences work can also trigger this category.¶

- `"frontier_llm"`¶

The request could assist the development of competing AI models, which is restricted under [Anthropic's commercial terms](https://www.anthropic.com/legal/commercial-terms). Benign machine learning work can also trigger this category.¶

- `"reasoning_extraction"`¶

The request asks the model to reproduce its internal reasoning in the response text. To get reasoning in a structured form instead, use [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/adaptive-thinking).¶

- `"general_harms"`¶

The request could be related to an area that was determined as harmful. Benign work might sometimes trigger this category.¶

- `explanation: string or null`¶

Human-readable explanation of the refusal.¶

This text is not guaranteed to be stable. `null` when no explanation is available for the category.¶

- `stop_reason: StopReason or null`¶

The reason that we stopped.¶

This may be one the following values:¶

* `"end_turn"`: the model reached a natural stopping point¶
* `"max_tokens"`: we exceeded the requested `max_tokens` or the model's maximum¶
* `"stop_sequence"`: one of your provided custom `stop_sequences` was generated¶
* `"tool_use"`: the model invoked one or more tools¶
* `"pause_turn"`: we paused a long-running turn. You may provide the response back as-is in a subsequent request to let the model continue.¶
* `"refusal"`: when streaming classifiers intervene to handle potential policy violations¶
* `"model_context_window_exceeded"`: we exceeded the model's context window¶

In non-streaming mode this value is always non-null. In streaming mode, it is null in the `message_start` event and non-null otherwise.¶

- `"end_turn"`¶

- `"max_tokens"`¶

- `"stop_sequence"`¶

- `"tool_use"`¶

- `"pause_turn"`¶

- `"refusal"`¶

- `"model_context_window_exceeded"`¶

- `stop_sequence: string or null`¶

Which custom stop sequence was generated, if any.¶

This value will be a non-null string if one of your custom stop sequences was generated.¶

- `usage: Usage`¶

Billing and rate-limit usage.¶

Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.¶

Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.¶

For example, `output_tokens` will be non-zero, even for an empty string response from Claude.¶

Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.¶

- `cache_creation: CacheCreation or null`¶

Breakdown of cached tokens by TTL¶

- `ephemeral_1h_input_tokens: number`¶

The number of input tokens used to create the 1 hour cache entry.¶

default: 0, minimum: 0¶

- `ephemeral_5m_input_tokens: number`¶

The number of input tokens used to create the 5 minute cache entry.¶

default: 0, minimum: 0¶

- `cache_creation_input_tokens: number or null`¶

The number of input tokens used to create the cache entry.¶

minimum: 0¶

- `cache_read_input_tokens: number or null`¶

The number of input tokens read from the cache.¶

minimum: 0¶

- `inference_geo: string or null`¶

The geographic region where inference was performed for this request.¶

- `input_tokens: number`¶

The number of input tokens which were used.¶

minimum: 0¶

- `output_tokens: number`¶

The number of output tokens which were used.¶

minimum: 0¶

- `output_tokens_details: OutputTokensDetails or null`¶

Breakdown of output tokens by category.¶

`output_tokens` remains the inclusive, authoritative total used for billing.¶
This object provides a read-only decomposition for observability — for example,¶
how many of the billed output tokens were spent on internal reasoning that may¶
have been summarized before being returned to you.¶

- `thinking_tokens: number`¶

Number of output tokens the model generated as internal reasoning, including¶
the thinking-block delimiter tokens.¶

Reflects the raw reasoning the model produced, not the (possibly shorter)¶
summarized thinking text returned in the response body. Computed by¶
re-tokenizing the raw reasoning text, so it may differ from the model's exact¶
generation count by a small number of tokens. Always ≤ `output_tokens`;¶
`output_tokens - thinking_tokens` approximates the non-reasoning output.¶

default: 0, minimum: 0¶

- `server_tool_use: ServerToolUsage or null`¶

The number of server tool requests.¶

- `web_fetch_requests: number`¶

The number of web fetch tool requests.¶

default: 0, minimum: 0¶

- `web_search_requests: number`¶

The number of web search tool requests.¶

default: 0, minimum: 0¶

- `service_tier: "standard" or "priority" or "batch" or null`¶

If the request used the priority, standard, or batch tier.¶

- `"standard"`¶

- `"priority"`¶

- `"batch"`¶

- `RawMessageDeltaEvent object`¶

- `type: "message_delta"`¶

default: message_delta¶

- `delta: object`¶

- `container: Container or null`¶

Information about the container used in the request (for the code execution tool)¶

- `stop_details: RefusalStopDetails or null`¶

Structured information about a refusal.¶

- `stop_reason: StopReason or null`¶

- `stop_sequence: string or null`¶

- `usage: MessageDeltaUsage`¶

Billing and rate-limit usage.¶

Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.¶

Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.¶

For example, `output_tokens` will be non-zero, even for an empty string response from Claude.¶

Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.¶

- `cache_creation_input_tokens: number or null`¶

The cumulative number of input tokens used to create the cache entry.¶

minimum: 0¶

- `cache_read_input_tokens: number or null`¶

The cumulative number of input tokens read from the cache.¶

minimum: 0¶

- `input_tokens: number or null`¶

The cumulative number of input tokens which were used.¶

minimum: 0¶

- `output_tokens: number`¶

The cumulative number of output tokens which were used.¶

- `output_tokens_details: OutputTokensDetails or null`¶

Breakdown of output tokens by category.¶

`output_tokens` remains the inclusive, authoritative total used for billing.¶
This object provides a read-only decomposition for observability — for example,¶
how many of the billed output tokens were spent on internal reasoning that may¶
have been summarized before being returned to you.¶

- `server_tool_use: ServerToolUsage or null`¶

The number of server tool requests.¶

- `RawMessageStopEvent object`¶

- `type: "message_stop"`¶

default: message_stop¶

- `RawContentBlockStartEvent object`¶

- `type: "content_block_start"`¶

default: content_block_start¶

- `content_block: TextBlock or ThinkingBlock or RedactedThinkingBlock or 9 more`¶

- `TextBlock object`¶

- `ThinkingBlock object`¶

- `RedactedThinkingBlock object`¶

- `ToolUseBlock object`¶

- `ServerToolUseBlock object`¶

- `WebSearchToolResultBlock object`¶

- `WebFetchToolResultBlock object`¶

- `CodeExecutionToolResultBlock object`¶

- `BashCodeExecutionToolResultBlock object`¶

- `TextEditorCodeExecutionToolResultBlock object`¶

- `ToolSearchToolResultBlock object`¶

- `ContainerUploadBlock object`¶

Response model for a file uploaded to the container.¶

- `index: number`¶

- `RawContentBlockDeltaEvent object`¶

- `type: "content_block_delta"`¶

default: content_block_delta¶

- `delta: RawContentBlockDelta`¶

- `TextDelta object`¶

- `type: "text_delta"`¶

default: text_delta¶

- `text: string`¶

- `InputJSONDelta object`¶

- `type: "input_json_delta"`¶

default: input_json_delta¶

- `partial_json: string`¶

- `CitationsDelta object`¶

- `type: "citations_delta"`¶

default: citations_delta¶

- `citation: CitationCharLocation or CitationPageLocation or CitationContentBlockLocation or 2 more`¶

- `CitationCharLocation object`¶

- `CitationPageLocation object`¶

- `CitationContentBlockLocation object`¶

- `CitationsWebSearchResultLocation object`¶

- `CitationsSearchResultLocation object`¶

- `ThinkingDelta object`¶

- `type: "thinking_delta"`¶

default: thinking_delta¶

- `thinking: string`¶

The incremental `thinking` text for this content block. Concatenate the `thinking` values of successive `thinking_delta` events to assemble the block's full `thinking` value.¶

- `SignatureDelta object`¶

- `type: "signature_delta"`¶

default: signature_delta¶

- `signature: string`¶

The `signature` for this thinking block: an opaque value used to verify that the block was generated by Claude when it is passed back to the API. Delivered in a `signature_delta` event just before the block's `content_block_stop` event.¶

- `index: number`¶

- `RawContentBlockStopEvent object`¶

- `type: "content_block_stop"`¶

default: content_block_stop¶

- `index: number`¶

### Redacted Thinking Block¶

- `RedactedThinkingBlock object`¶

- `type: "redacted_thinking"`¶

default: redacted_thinking¶

- `data: string`¶

The contents of this redacted thinking block, returned when portions of the model's thinking were safety-redacted. This field is opaque and encrypted, with no readable content.¶

Pass `redacted_thinking` blocks back to the API unchanged when continuing a multi-turn conversation.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking#redacted-thinking-blocks) for details.¶

### Redacted Thinking Block Param¶

- `RedactedThinkingBlockParam object`¶

- `type: "redacted_thinking"`¶

- `data: string`¶

The `data` value of this redacted thinking block, exactly as returned by the API in a previous response. Opaque and encrypted; pass it back unchanged.¶

### Refusal Stop Details¶

- `RefusalStopDetails object`¶

Structured information about a refusal.¶

- `type: "refusal"`¶

default: refusal¶

- `category: "cyber" or "bio" or "frontier_llm" or 2 more or null`¶

The policy category that triggered a refusal.¶

- `"cyber"`¶

The request could enable cyber harm, such as malware or exploit development. Benign cybersecurity work can also trigger this category.¶

- `"bio"`¶

The request could enable biological harm, such as dangerous lab methods. Beneficial life sciences work can also trigger this category.¶

- `"frontier_llm"`¶

The request could assist the development of competing AI models, which is restricted under [Anthropic's commercial terms](https://www.anthropic.com/legal/commercial-terms). Benign machine learning work can also trigger this category.¶

- `"reasoning_extraction"`¶

The request asks the model to reproduce its internal reasoning in the response text. To get reasoning in a structured form instead, use [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/adaptive-thinking).¶

- `"general_harms"`¶

The request could be related to an area that was determined as harmful. Benign work might sometimes trigger this category.¶

- `explanation: string or null`¶

Human-readable explanation of the refusal.¶

This text is not guaranteed to be stable. `null` when no explanation is available for the category.¶

### Search Result Block Param¶

- `SearchResultBlockParam object`¶

- `type: "search_result"`¶

- `content: array of TextBlockParam`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `source: string`¶

- `title: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam`¶

- `enabled: optional boolean`¶

### Server Tool Caller¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Server Tool Caller 20260120¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Server Tool Usage¶

- `ServerToolUsage object`¶

- `web_fetch_requests: number`¶

The number of web fetch tool requests.¶

default: 0, minimum: 0¶

- `web_search_requests: number`¶

The number of web search tool requests.¶

default: 0, minimum: 0¶

### Server Tool Use Block¶

- `ServerToolUseBlock object`¶

- `type: "server_tool_use"`¶

default: server_tool_use¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

### Server Tool Use Block Param¶

- `ServerToolUseBlockParam object`¶

- `type: "server_tool_use"`¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Signature Delta¶

- `SignatureDelta object`¶

- `type: "signature_delta"`¶

default: signature_delta¶

- `signature: string`¶

The `signature` for this thinking block: an opaque value used to verify that the block was generated by Claude when it is passed back to the API. Delivered in a `signature_delta` event just before the block's `content_block_stop` event.¶

### Skill Params¶

- `SkillParams object`¶

Specification for a skill to be loaded in a container (request model).¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: optional string`¶

Skill version or 'latest' for most recent version¶

maxLength: 64, minLength: 1¶

### Stop Reason¶

- `StopReason = "end_turn" or "max_tokens" or "stop_sequence" or 4 more`¶

- `"end_turn"`¶

- `"max_tokens"`¶

- `"stop_sequence"`¶

- `"tool_use"`¶

- `"pause_turn"`¶

- `"refusal"`¶

- `"model_context_window_exceeded"`¶

### Text Block¶

- `TextBlock object`¶

- `type: "text"`¶

default: text¶

- `citations: array of TextCitation or null`¶

Citations supporting the text block.¶

The type of citation returned will depend on the type of document being cited. Citing a PDF results in `page_location`, plain text results in `char_location`, and content document results in `content_block_location`.¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `text: string`¶

minLength: 0¶

### Text Block Param¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

### Text Citation¶

- `TextCitation = CitationCharLocation or CitationPageLocation or CitationContentBlockLocation or 2 more`¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

### Text Citation Param¶

- `TextCitationParam = CitationCharLocationParam or CitationPageLocationParam or CitationContentBlockLocationParam or 2 more`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

### Text Delta¶

- `TextDelta object`¶

- `type: "text_delta"`¶

default: text_delta¶

- `text: string`¶

### Text Editor Code Execution Create Result Block¶

- `TextEditorCodeExecutionCreateResultBlock object`¶

- `type: "text_editor_code_execution_create_result"`¶

default: text_editor_code_execution_create_result¶

- `is_file_update: boolean`¶

### Text Editor Code Execution Create Result Block Param¶

- `TextEditorCodeExecutionCreateResultBlockParam object`¶

- `type: "text_editor_code_execution_create_result"`¶

- `is_file_update: boolean`¶

### Text Editor Code Execution Str Replace Result Block¶

- `TextEditorCodeExecutionStrReplaceResultBlock object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

default: text_editor_code_execution_str_replace_result¶

- `lines: array of string or null`¶

- `new_lines: number or null`¶

- `new_start: number or null`¶

- `old_lines: number or null`¶

- `old_start: number or null`¶

### Text Editor Code Execution Str Replace Result Block Param¶

- `TextEditorCodeExecutionStrReplaceResultBlockParam object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

- `lines: optional array of string or null`¶

- `new_lines: optional number or null`¶

- `new_start: optional number or null`¶

- `old_lines: optional number or null`¶

- `old_start: optional number or null`¶

### Text Editor Code Execution Tool Result Block¶

- `TextEditorCodeExecutionToolResultBlock object`¶

- `type: "text_editor_code_execution_tool_result"`¶

default: text_editor_code_execution_tool_result¶

- `content: TextEditorCodeExecutionToolResultError or TextEditorCodeExecutionViewResultBlock or TextEditorCodeExecutionCreateResultBlock or TextEditorCodeExecutionStrReplaceResultBlock`¶

- `TextEditorCodeExecutionToolResultError object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

default: text_editor_code_execution_tool_result_error¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: string or null`¶

- `TextEditorCodeExecutionViewResultBlock object`¶

- `type: "text_editor_code_execution_view_result"`¶

default: text_editor_code_execution_view_result¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: number or null`¶

- `start_line: number or null`¶

- `total_lines: number or null`¶

- `TextEditorCodeExecutionCreateResultBlock object`¶

- `type: "text_editor_code_execution_create_result"`¶

default: text_editor_code_execution_create_result¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlock object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

default: text_editor_code_execution_str_replace_result¶

- `lines: array of string or null`¶

- `new_lines: number or null`¶

- `new_start: number or null`¶

- `old_lines: number or null`¶

- `old_start: number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Text Editor Code Execution Tool Result Block Param¶

- `TextEditorCodeExecutionToolResultBlockParam object`¶

- `type: "text_editor_code_execution_tool_result"`¶

- `content: TextEditorCodeExecutionToolResultErrorParam or TextEditorCodeExecutionViewResultBlockParam or TextEditorCodeExecutionCreateResultBlockParam or TextEditorCodeExecutionStrReplaceResultBlockParam`¶

- `TextEditorCodeExecutionToolResultErrorParam object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: optional string or null`¶

- `TextEditorCodeExecutionViewResultBlockParam object`¶

- `type: "text_editor_code_execution_view_result"`¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: optional number or null`¶

- `start_line: optional number or null`¶

- `total_lines: optional number or null`¶

- `TextEditorCodeExecutionCreateResultBlockParam object`¶

- `type: "text_editor_code_execution_create_result"`¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlockParam object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

- `lines: optional array of string or null`¶

- `new_lines: optional number or null`¶

- `new_start: optional number or null`¶

- `old_lines: optional number or null`¶

- `old_start: optional number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

### Text Editor Code Execution Tool Result Error¶

- `TextEditorCodeExecutionToolResultError object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

default: text_editor_code_execution_tool_result_error¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: string or null`¶

### Text Editor Code Execution Tool Result Error Code¶

- `TextEditorCodeExecutionToolResultErrorCode = "invalid_tool_input" or "unavailable" or "too_many_requests" or 2 more`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

### Text Editor Code Execution Tool Result Error Param¶

- `TextEditorCodeExecutionToolResultErrorParam object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: optional string or null`¶

### Text Editor Code Execution View Result Block¶

- `TextEditorCodeExecutionViewResultBlock object`¶

- `type: "text_editor_code_execution_view_result"`¶

default: text_editor_code_execution_view_result¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: number or null`¶

- `start_line: number or null`¶

- `total_lines: number or null`¶

### Text Editor Code Execution View Result Block Param¶

- `TextEditorCodeExecutionViewResultBlockParam object`¶

- `type: "text_editor_code_execution_view_result"`¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: optional number or null`¶

- `start_line: optional number or null`¶

- `total_lines: optional number or null`¶

### Thinking Block¶

- `ThinkingBlock object`¶

- `type: "thinking"`¶

default: thinking¶

- `signature: string`¶

A value used to verify that this thinking block was generated by Claude when it is passed back to the API.¶

This is an opaque field and should not be interpreted or parsed. When passing thinking blocks back to the API (required when using tools with extended thinking), pass them back exactly as received, with this field intact.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `thinking: string`¶

The text of Claude's thinking process for this block.¶

### Thinking Block Param¶

- `ThinkingBlockParam object`¶

- `type: "thinking"`¶

- `signature: string`¶

The `signature` value of this thinking block, exactly as returned by the API in a previous response. Used to verify that the block was generated by Claude.¶

Thinking blocks must be passed back unmodified and in their original order; a modified block results in a 400 `invalid_request_error`.¶

- `thinking: string`¶

The `thinking` text of this block as returned by the API.¶

### Thinking Config Adaptive¶

- `ThinkingConfigAdaptive object`¶

- `type: "adaptive"`¶

- `display: optional "summarized" or "omitted" or null`¶

Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.¶

- `"summarized"`¶

- `"omitted"`¶

### Thinking Config Disabled¶

- `ThinkingConfigDisabled object`¶

- `type: "disabled"`¶

### Thinking Config Enabled¶

- `ThinkingConfigEnabled object`¶

- `type: "enabled"`¶

- `budget_tokens: number`¶

Determines how many tokens Claude can use for its internal reasoning process. Larger budgets can enable more thorough analysis for complex problems, improving response quality.¶

Must be ≥1024 and less than `max_tokens`.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

minimum: 1024¶

- `display: optional "summarized" or "omitted" or null`¶

Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.¶

- `"summarized"`¶

- `"omitted"`¶

### Thinking Config Param¶

- `ThinkingConfigParam = ThinkingConfigEnabled or ThinkingConfigDisabled or ThinkingConfigAdaptive`¶

Configuration for enabling Claude's extended thinking.¶

When enabled, responses include `thinking` content blocks showing Claude's thinking process before the final answer. Requires a minimum budget of 1,024 tokens and counts towards your `max_tokens` limit.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `ThinkingConfigEnabled object`¶

- `type: "enabled"`¶

- `budget_tokens: number`¶

Determines how many tokens Claude can use for its internal reasoning process. Larger budgets can enable more thorough analysis for complex problems, improving response quality.¶

Must be ≥1024 and less than `max_tokens`.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

minimum: 1024¶

- `display: optional "summarized" or "omitted" or null`¶

Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.¶

- `"summarized"`¶

- `"omitted"`¶

- `ThinkingConfigDisabled object`¶

- `type: "disabled"`¶

- `ThinkingConfigAdaptive object`¶

- `type: "adaptive"`¶

- `display: optional "summarized" or "omitted" or null`¶

Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.¶

- `"summarized"`¶

- `"omitted"`¶

### Thinking Delta¶

- `ThinkingDelta object`¶

- `type: "thinking_delta"`¶

default: thinking_delta¶

- `thinking: string`¶

The incremental `thinking` text for this content block. Concatenate the `thinking` values of successive `thinking_delta` events to assemble the block's full `thinking` value.¶

### Tool¶

- `Tool object`¶

- `type: optional "custom" or null`¶

- `input_schema: object`¶

[JSON schema](https://json-schema.org/draft/2020-12) for this tool's input.¶

This defines the shape of the `input` that your tool accepts and that the model will produce.¶

- `type: "object"`¶

- `properties: optional map[unknown] or null`¶

- `required: optional array of string or null`¶

- `name: string`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

maxLength: 128, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,128}$¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `description: optional string`¶

Description of what this tool does.¶

Tool descriptions should be as detailed as possible. The more information that the model has about what the tool is and how to use it, the better it will perform. You can use natural language descriptions to reinforce important aspects of the tool input JSON schema.¶

- `eager_input_streaming: optional boolean or null`¶

Enable eager input streaming for this tool. When true, tool input parameters will be streamed incrementally as they are generated, and types will be inferred on-the-fly rather than buffering the full JSON output. When false, streaming is disabled for this tool even if the fine-grained-tool-streaming beta is active. When null (default), uses the default behavior based on beta headers.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Tool Bash 20250124¶

- `ToolBash20250124 object`¶

- `type: "bash_20250124"`¶

- `name: "bash"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Tool Choice¶

- `ToolChoice = ToolChoiceAuto or ToolChoiceAny or ToolChoiceTool or ToolChoiceNone`¶

How the model should use the provided tools. The model can use a specific tool, any available tool, decide by itself, or not use tools at all.¶

- `ToolChoiceAuto object`¶

The model will automatically decide whether to use tools.¶

- `type: "auto"`¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output at most one tool use.¶

- `ToolChoiceAny object`¶

The model will use any available tools.¶

- `type: "any"`¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output exactly one tool use.¶

- `ToolChoiceTool object`¶

The model will use the specified tool with `tool_choice.name`.¶

- `type: "tool"`¶

- `name: string`¶

The name of the tool to use.¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output exactly one tool use.¶

- `ToolChoiceNone object`¶

The model will not be allowed to use tools.¶

- `type: "none"`¶

### Tool Choice Any¶

- `ToolChoiceAny object`¶

The model will use any available tools.¶

- `type: "any"`¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output exactly one tool use.¶

### Tool Choice Auto¶

- `ToolChoiceAuto object`¶

The model will automatically decide whether to use tools.¶

- `type: "auto"`¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output at most one tool use.¶

### Tool Choice None¶

- `ToolChoiceNone object`¶

The model will not be allowed to use tools.¶

- `type: "none"`¶

### Tool Choice Tool¶

- `ToolChoiceTool object`¶

The model will use the specified tool with `tool_choice.name`.¶

- `type: "tool"`¶

- `name: string`¶

The name of the tool to use.¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output exactly one tool use.¶

### Tool Reference Block¶

- `ToolReferenceBlock object`¶

- `type: "tool_reference"`¶

default: tool_reference¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

### Tool Reference Block Param¶

- `ToolReferenceBlockParam object`¶

Tool reference block that can be included in tool_result content.¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

### Tool Result Block Param¶

- `ToolResultBlockParam object`¶

- `type: "tool_result"`¶

- `tool_use_id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `content: optional string or array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `string`¶

- `array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

- `SearchResultBlockParam object`¶

- `type: "search_result"`¶

- `content: array of TextBlockParam`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional array of TextCitationParam or null`¶

- `source: string`¶

- `title: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam`¶

- `enabled: optional boolean`¶

- `DocumentBlockParam object`¶

- `type: "document"`¶

- `source: Base64PDFSource or PlainTextSource or ContentBlockSource or 2 more`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `ContentBlockSource object`¶

- `type: "content"`¶

- `content: string or array of ContentBlockSourceContent`¶

- `string`¶

- `ContentBlockSourceContent = array of ContentBlockSourceContent`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `URLPDFSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileDocumentSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

- `context: optional string or null`¶

minLength: 1¶

- `title: optional string or null`¶

maxLength: 500, minLength: 1¶

- `ToolReferenceBlockParam object`¶

Tool reference block that can be included in tool_result content.¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BrowserStateBlockParam object`¶

The caller's browser state after a browser toolset member call —¶
the full inventory of open tabs, which tab is active, and any side¶
effects (tabs opened, download state changes) the call produced.¶

At most one per `tool_result`, only on a non-error result answering a¶
browser toolset member `tool_use`. The server renders the¶
model-visible text from it; the model never sees the raw fields.¶

- `type: "browser_state"`¶

- `tabs: array of BrowserStateTabEntry`¶

All tabs open in the browser after this call — the full inventory, not a delta. May be empty. Whenever non-empty, exactly one entry carries `active: true`.¶

maxItems: 100¶

- `tab_id: string`¶

The caller-assigned identifier for this tab, unique within the inventory.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `title: string`¶

The title of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The URL of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `active: optional boolean`¶

Whether this tab is the active tab after this call. Whenever `tabs` is non-empty, exactly one entry is marked `active: true`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `state_changes: optional array of BrowserStateChange or null`¶

Tabs opened and download state changes during this call. "Nothing to report" is expressed by omitting the field, never by an empty list.¶

maxItems: 200, minItems: 1¶

- `BrowserStateChangeTabOpened object`¶

A tab this call's execution opened that remains open at its end —¶
the creation delta of the `tabs` inventory, not an event log.¶

Carries only the `tab_id`; the tab's `title` and `url` live on its¶
`tabs` entry, which must include the same `tab_id`. A tab opened¶
during a failed call gets no deferred `tab_opened`; it simply appears¶
in the next result's `tabs` inventory.¶

- `type: "tab_opened"`¶

- `tab_id: string`¶

The `tab_id` of the opened tab, present in `tabs`.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadStarted object`¶

A file download that started during this call.¶

- `type: "download_started"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadCompleted object`¶

A file download that finished during this call, reported with the¶
same `download_id` as its `download_started` — or without a prior¶
`download_started`, when the download finished during the call that¶
started it (at most one state change per `download_id` per result).¶

- `type: "download_completed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `path: optional string or null`¶

Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `size_bytes: optional number or null`¶

The completed download's size.¶

minimum: 0¶

- `BrowserStateChangeDownloadFailed object`¶

A file download that failed — or was cancelled — during this call.¶

- `type: "download_failed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `error: optional string or null`¶

The failure or cancellation detail, when known.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `is_error: optional boolean`¶

- `toolset_name: optional string or null`¶

For a toolset member tool_result, the toolset family of the paired tool_use.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

### Tool Search Tool Bm25 20251119¶

- `ToolSearchToolBm25_20251119 object`¶

- `type: "tool_search_tool_bm25_20251119" or "tool_search_tool_bm25"`¶

- `"tool_search_tool_bm25_20251119"`¶

- `"tool_search_tool_bm25"`¶

- `name: "tool_search_tool_bm25"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Tool Search Tool Regex 20251119¶

- `ToolSearchToolRegex20251119 object`¶

- `type: "tool_search_tool_regex_20251119" or "tool_search_tool_regex"`¶

- `"tool_search_tool_regex_20251119"`¶

- `"tool_search_tool_regex"`¶

- `name: "tool_search_tool_regex"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Tool Search Tool Result Block¶

- `ToolSearchToolResultBlock object`¶

- `type: "tool_search_tool_result"`¶

default: tool_search_tool_result¶

- `content: ToolSearchToolResultError or ToolSearchToolSearchResultBlock`¶

- `ToolSearchToolResultError object`¶

- `type: "tool_search_tool_result_error"`¶

default: tool_search_tool_result_error¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: string or null`¶

- `ToolSearchToolSearchResultBlock object`¶

- `type: "tool_search_tool_search_result"`¶

default: tool_search_tool_search_result¶

- `tool_references: array of ToolReferenceBlock`¶

- `type: "tool_reference"`¶

default: tool_reference¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Tool Search Tool Result Block Param¶

- `ToolSearchToolResultBlockParam object`¶

- `type: "tool_search_tool_result"`¶

- `content: ToolSearchToolResultErrorParam or ToolSearchToolSearchResultBlockParam`¶

- `ToolSearchToolResultErrorParam object`¶

- `type: "tool_search_tool_result_error"`¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: optional string or null`¶

- `ToolSearchToolSearchResultBlockParam object`¶

- `type: "tool_search_tool_search_result"`¶

- `tool_references: array of ToolReferenceBlockParam`¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

### Tool Search Tool Result Error¶

- `ToolSearchToolResultError object`¶

- `type: "tool_search_tool_result_error"`¶

default: tool_search_tool_result_error¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: string or null`¶

### Tool Search Tool Result Error Code¶

- `ToolSearchToolResultErrorCode = "invalid_tool_input" or "unavailable" or "too_many_requests" or "execution_time_exceeded"`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

### Tool Search Tool Result Error Param¶

- `ToolSearchToolResultErrorParam object`¶

- `type: "tool_search_tool_result_error"`¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: optional string or null`¶

### Tool Search Tool Search Result Block¶

- `ToolSearchToolSearchResultBlock object`¶

- `type: "tool_search_tool_search_result"`¶

default: tool_search_tool_search_result¶

- `tool_references: array of ToolReferenceBlock`¶

- `type: "tool_reference"`¶

default: tool_reference¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

### Tool Search Tool Search Result Block Param¶

- `ToolSearchToolSearchResultBlockParam object`¶

- `type: "tool_search_tool_search_result"`¶

- `tool_references: array of ToolReferenceBlockParam`¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

### Tool Text Editor 20250124¶

- `ToolTextEditor20250124 object`¶

- `type: "text_editor_20250124"`¶

- `name: "str_replace_editor"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Tool Text Editor 20250429¶

- `ToolTextEditor20250429 object`¶

- `type: "text_editor_20250429"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Tool Text Editor 20250728¶

- `ToolTextEditor20250728 object`¶

- `type: "text_editor_20250728"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `max_characters: optional number or null`¶

Maximum number of characters to display when viewing a file. If not specified, defaults to displaying the full file.¶

minimum: 1¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Tool Union¶

- `ToolUnion = Tool or ToolBash20250124 or CodeExecutionTool20250522 or 18 more`¶

- `Tool object`¶

- `type: optional "custom" or null`¶

- `input_schema: object`¶

[JSON schema](https://json-schema.org/draft/2020-12) for this tool's input.¶

This defines the shape of the `input` that your tool accepts and that the model will produce.¶

- `type: "object"`¶

- `properties: optional map[unknown] or null`¶

- `required: optional array of string or null`¶

- `name: string`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

maxLength: 128, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,128}$¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `description: optional string`¶

Description of what this tool does.¶

Tool descriptions should be as detailed as possible. The more information that the model has about what the tool is and how to use it, the better it will perform. You can use natural language descriptions to reinforce important aspects of the tool input JSON schema.¶

- `eager_input_streaming: optional boolean or null`¶

Enable eager input streaming for this tool. When true, tool input parameters will be streamed incrementally as they are generated, and types will be inferred on-the-fly rather than buffering the full JSON output. When false, streaming is disabled for this tool even if the fine-grained-tool-streaming beta is active. When null (default), uses the default behavior based on beta headers.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolBash20250124 object`¶

- `type: "bash_20250124"`¶

- `name: "bash"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20250522 object`¶

- `type: "code_execution_20250522"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20250825 object`¶

- `type: "code_execution_20250825"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20260120 object`¶

Code execution tool with REPL state persistence (daemon mode + gVisor checkpoint).¶

- `type: "code_execution_20260120"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20260521 object`¶

Code execution tool with REPL state persistence.¶

- `type: "code_execution_20260521"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `BrowserToolset20260801 object`¶

The browser toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the browser tool family. The model is served¶
the family's tool with any members disabled via `configs` removed¶
from its schema.¶

- `type: "browser_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `configs: optional BrowserToolsetConfigs or null`¶

Per-member configuration for `browser_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional BrowserTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `close_tab: optional BrowserCloseTabConfig or null`¶

`close_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional BrowserDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `file_upload: optional BrowserFileUploadConfig or null`¶

`file_upload`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `find: optional BrowserFindConfig or null`¶

`find`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `form_input: optional BrowserFormInputConfig or null`¶

`form_input`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `get_page_text: optional BrowserGetPageTextConfig or null`¶

`get_page_text`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional BrowserHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hover: optional BrowserHoverConfig or null`¶

`hover`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `javascript_exec: optional BrowserJavascriptExecConfig or null`¶

`javascript_exec`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional BrowserKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional BrowserLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional BrowserLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional BrowserLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional BrowserLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `list_tabs: optional BrowserListTabsConfig or null`¶

`list_tabs`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional BrowserMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional BrowserMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `navigate: optional BrowserNavigateConfig or null`¶

`navigate`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `new_tab: optional BrowserNewTabConfig or null`¶

`new_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_console: optional BrowserReadConsoleConfig or null`¶

`read_console`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_network: optional BrowserReadNetworkConfig or null`¶

`read_network`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_page: optional BrowserReadPageConfig or null`¶

`read_page`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional BrowserRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional BrowserScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional BrowserScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll_to: optional BrowserScrollToConfig or null`¶

`scroll_to`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `switch_tab: optional BrowserSwitchTabConfig or null`¶

`switch_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional BrowserTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional BrowserWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional BrowserZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `MemoryTool20250818 object`¶

- `type: "memory_20250818"`¶

- `name: "memory"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ComputerToolset20260801 object`¶

The computer toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the computer tool family. The model is¶
served the family's tool with any members disabled via `configs`¶
removed from its schema. Every member is enabled by default, zoom¶
included. The single-tool options `display_number` and¶
`enable_zoom` are not fields of a toolset entry — it carries only¶
`type`, `configs`, and `cache_control`; zoom is controlled¶
via `configs.zoom.enabled`.¶

- `type: "computer_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `configs: optional ComputerToolsetConfigs or null`¶

Per-member configuration for `computer_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional ComputerTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `cursor_position: optional ComputerCursorPositionConfig or null`¶

`cursor_position`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional ComputerDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional ComputerHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional ComputerKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional ComputerLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional ComputerLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional ComputerLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional ComputerLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional ComputerMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional ComputerMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional ComputerRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional ComputerScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional ComputerScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional ComputerTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional ComputerWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional ComputerZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `ToolTextEditor20250124 object`¶

- `type: "text_editor_20250124"`¶

- `name: "str_replace_editor"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolTextEditor20250429 object`¶

- `type: "text_editor_20250429"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolTextEditor20250728 object`¶

- `type: "text_editor_20250728"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `max_characters: optional number or null`¶

Maximum number of characters to display when viewing a file. If not specified, defaults to displaying the full file.¶

minimum: 1¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `WebSearchTool20250305 object`¶

- `type: "web_search_20250305"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `type: "approximate"`¶

- `city: optional string or null`¶

The city of the user.¶

maxLength: 255, minLength: 1¶

- `country: optional string or null`¶

The two letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2) of the user.¶

maxLength: 2, minLength: 2¶

- `region: optional string or null`¶

The region of the user.¶

maxLength: 255, minLength: 1¶

- `timezone: optional string or null`¶

The [IANA timezone](https://nodatime.org/TimeZones) of the user.¶

maxLength: 255, minLength: 1¶

- `WebFetchTool20250910 object`¶

- `type: "web_fetch_20250910"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `enabled: optional boolean`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

- `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`¶

Whether URLs in user messages are fetchable: "all" or "none".¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

 - `WebSearchTool20260209 object`¶

- `type: "web_search_20260209"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `WebFetchTool20260209 object`¶

- `type: "web_fetch_20260209"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `WebFetchTool20260309 object`¶

Web fetch tool with use_cache parameter for bypassing cached content.¶

- `type: "web_fetch_20260309"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

- `WebSearchTool20260318 object`¶

- `type: "web_search_20260318"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `WebFetchTool20260318 object`¶

- `type: "web_fetch_20260318"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

- `ToolSearchToolBm25_20251119 object`¶

- `type: "tool_search_tool_bm25_20251119" or "tool_search_tool_bm25"`¶

- `"tool_search_tool_bm25_20251119"`¶

- `"tool_search_tool_bm25"`¶

- `name: "tool_search_tool_bm25"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolSearchToolRegex20251119 object`¶

- `type: "tool_search_tool_regex_20251119" or "tool_search_tool_regex"`¶

- `"tool_search_tool_regex_20251119"`¶

- `"tool_search_tool_regex"`¶

- `name: "tool_search_tool_regex"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Tool Use Block¶

- `ToolUseBlock object`¶

- `type: "tool_use"`¶

default: tool_use¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: string`¶

minLength: 1¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

### Tool Use Block Param¶

- `ToolUseBlockParam object`¶

- `type: "tool_use"`¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `input: map[unknown]`¶

- `name: string`¶

maxLength: 200, minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family this member belongs to.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

### URL Image Source¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

### URL PDF Source¶

- `URLPDFSource object`¶

- `type: "url"`¶

- `url: string`¶

### Usage¶

- `Usage object`¶

- `cache_creation: CacheCreation or null`¶

Breakdown of cached tokens by TTL¶

- `ephemeral_1h_input_tokens: number`¶

The number of input tokens used to create the 1 hour cache entry.¶

default: 0, minimum: 0¶

- `ephemeral_5m_input_tokens: number`¶

The number of input tokens used to create the 5 minute cache entry.¶

default: 0, minimum: 0¶

- `cache_creation_input_tokens: number or null`¶

The number of input tokens used to create the cache entry.¶

minimum: 0¶

- `cache_read_input_tokens: number or null`¶

The number of input tokens read from the cache.¶

minimum: 0¶

- `inference_geo: string or null`¶

The geographic region where inference was performed for this request.¶

- `input_tokens: number`¶

The number of input tokens which were used.¶

minimum: 0¶

- `output_tokens: number`¶

The number of output tokens which were used.¶

minimum: 0¶

- `output_tokens_details: OutputTokensDetails or null`¶

Breakdown of output tokens by category.¶

`output_tokens` remains the inclusive, authoritative total used for billing.¶
This object provides a read-only decomposition for observability — for example,¶
how many of the billed output tokens were spent on internal reasoning that may¶
have been summarized before being returned to you.¶

- `thinking_tokens: number`¶

Number of output tokens the model generated as internal reasoning, including¶
the thinking-block delimiter tokens.¶

Reflects the raw reasoning the model produced, not the (possibly shorter)¶
summarized thinking text returned in the response body. Computed by¶
re-tokenizing the raw reasoning text, so it may differ from the model's exact¶
generation count by a small number of tokens. Always ≤ `output_tokens`;¶
`output_tokens - thinking_tokens` approximates the non-reasoning output.¶

default: 0, minimum: 0¶

- `server_tool_use: ServerToolUsage or null`¶

The number of server tool requests.¶

- `web_fetch_requests: number`¶

The number of web fetch tool requests.¶

default: 0, minimum: 0¶

- `web_search_requests: number`¶

The number of web search tool requests.¶

default: 0, minimum: 0¶

- `service_tier: "standard" or "priority" or "batch" or null`¶

If the request used the priority, standard, or batch tier.¶

- `"standard"`¶

- `"priority"`¶

- `"batch"`¶

### User Location¶

- `UserLocation object`¶

- `type: "approximate"`¶

- `city: optional string or null`¶

The city of the user.¶

maxLength: 255, minLength: 1¶

- `country: optional string or null`¶

The two letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2) of the user.¶

maxLength: 2, minLength: 2¶

- `region: optional string or null`¶

The region of the user.¶

maxLength: 255, minLength: 1¶

- `timezone: optional string or null`¶

The [IANA timezone](https://nodatime.org/TimeZones) of the user.¶

maxLength: 255, minLength: 1¶

### Web Fetch Block¶

- `WebFetchBlock object`¶

- `type: "web_fetch_result"`¶

default: web_fetch_result¶

- `content: DocumentBlock`¶

- `type: "document"`¶

default: document¶

- `citations: CitationsConfig or null`¶

Citation configuration for the document¶

- `enabled: boolean`¶

default: false¶

- `source: Base64PDFSource or PlainTextSource`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `title: string or null`¶

The title of the document¶

- `retrieved_at: string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `url: string`¶

Fetched content URL¶

### Web Fetch Block Param¶

- `WebFetchBlockParam object`¶

- `type: "web_fetch_result"`¶

- `content: DocumentBlockParam`¶

- `type: "document"`¶

- `source: Base64PDFSource or PlainTextSource or ContentBlockSource or 2 more`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `ContentBlockSource object`¶

- `type: "content"`¶

- `content: string or array of ContentBlockSourceContent`¶

- `string`¶

- `ContentBlockSourceContent = array of ContentBlockSourceContent`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

- `URLPDFSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileDocumentSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

- `enabled: optional boolean`¶

- `context: optional string or null`¶

minLength: 1¶

- `title: optional string or null`¶

maxLength: 500, minLength: 1¶

- `url: string`¶

Fetched content URL¶

- `retrieved_at: optional string or null`¶

ISO 8601 timestamp when the content was retrieved¶

### Web Fetch Tool 20250910¶

- `WebFetchTool20250910 object`¶

- `type: "web_fetch_20250910"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `enabled: optional boolean`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Web Fetch Tool 20260209¶

- `WebFetchTool20260209 object`¶

- `type: "web_fetch_20260209"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `enabled: optional boolean`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

### Web Fetch Tool 20260309¶

- `WebFetchTool20260309 object`¶

Web fetch tool with use_cache parameter for bypassing cached content.¶

- `type: "web_fetch_20260309"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `enabled: optional boolean`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

### Web Fetch Tool 20260318¶

- `WebFetchTool20260318 object`¶

- `type: "web_fetch_20260318"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `enabled: optional boolean`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

- `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`¶

Whether URLs in user messages are fetchable: "all" or "none".¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

### Web Fetch Tool 20260209¶

- `WebFetchTool20260209 object`¶

- `type: "web_fetch_20260209"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `enabled: optional boolean`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

- `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`¶

Whether URLs in user messages are fetchable: "all" or "none".¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

### Web Fetch Tool 20260309¶

- `WebFetchTool20260309 object`¶

Web fetch tool with use_cache parameter for bypassing cached content.¶

- `type: "web_fetch_20260309"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `enabled: optional boolean`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

- `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`¶

Whether URLs in user messages are fetchable: "all" or "none".¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

### Web Fetch Tool 20260318¶

- `WebFetchTool20260318 object`¶

- `type: "web_fetch_20260318"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `enabled: optional boolean`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

- `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`¶

Whether URLs in user messages are fetchable: "all" or "none".¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.

- `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

### Web Fetch Tool Result Block¶

- `WebFetchToolResultBlock object`¶

- `type: "web_fetch_tool_result"`¶

default: web_fetch_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `content: WebFetchToolResultErrorBlock or WebFetchBlock`¶

- `WebFetchToolResultErrorBlock object`¶

- `type: "web_fetch_tool_result_error"`¶

default: web_fetch_tool_result_error¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlock object`¶

- `type: "web_fetch_result"`¶

default: web_fetch_result¶

- `content: DocumentBlock`¶

- `type: "document"`¶

default: document¶

- `citations: CitationsConfig or null`¶

Citation configuration for the document¶

- `enabled: boolean`¶

default: false¶

- `source: Base64PDFSource or PlainTextSource`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `title: string or null`¶

The title of the document¶

- `retrieved_at: string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `url: string`¶

Fetched content URL¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Web Fetch Tool Result Block Param¶

- `WebFetchToolResultBlockParam object`¶

- `type: "web_fetch_tool_result"`¶

- `content: WebFetchToolResultErrorBlockParam or WebFetchBlockParam`¶

- `WebFetchToolResultErrorBlockParam object`¶

- `type: "web_fetch_tool_result_error"`¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlockParam object`¶

- `type: "web_fetch_result"`¶

- `content: DocumentBlockParam`¶

- `type: "document"`¶

- `source: Base64PDFSource or PlainTextSource or ContentBlockSource or 2 more`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `ContentBlockSource object`¶

- `type: "content"`¶

- `content: string or array of ContentBlockSourceContent`¶

- `string`¶

- `ContentBlockSourceContent = array of ContentBlockSourceContent`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

- `URLPDFSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileDocumentSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

- `enabled: optional boolean`¶

- `context: optional string or null`¶

minLength: 1¶

- `title: optional string or null`¶

maxLength: 500, minLength: 1¶

- `url: string`¶

Fetched content URL¶

- `retrieved_at: optional string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Web Fetch Tool Result Error Block¶

- `WebFetchToolResultErrorBlock object`¶

- `type: "web_fetch_tool_result_error"`¶

default: web_fetch_tool_result_error¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

### Web Fetch Tool Result Error Block Param¶

- `WebFetchToolResultErrorBlockParam object`¶

- `type: "web_fetch_tool_result_error"`¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

### Web Fetch Tool Result Error Code¶

- `WebFetchToolResultErrorCode = "invalid_tool_input" or "url_too_long" or "url_not_allowed" or 7 more`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

### Web Fetch URL Source All¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

### Web Fetch URL Source Except¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

### Web Fetch URL Source None¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

### Web Fetch URL Source Only¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

### Web Fetch URL Source Tool Reference¶

- `WebFetchURLSourceToolReference object`¶

One entry of a tool filter's `tools`: it must name a tool declared¶
in this request's `tools[]`.¶

- `type: "tool_reference"`¶

- `name: string`¶

### Web Fetch URL Sources¶

- `WebFetchURLSources object`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

- `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`¶

Whether URLs in user messages are fetchable: "all" or "none".¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

### Web Search Result Block¶

- `WebSearchResultBlock object`¶

- `type: "web_search_result"`¶

default: web_search_result¶

- `encrypted_content: string`¶

- `page_age: string or null`¶

- `title: string`¶

- `url: string`¶

### Web Search Result Block Param¶

- `WebSearchResultBlockParam object`¶

- `type: "web_search_result"`¶

- `encrypted_content: string`¶

- `title: string`¶

- `url: string`¶

- `page_age: optional string or null`¶

### Web Search Tool 20250305¶

- `WebSearchTool20250305 object`¶

- `type: "web_search_20250305"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `type: "approximate"`¶

- `city: optional string or null`¶

The city of the user.¶

maxLength: 255, minLength: 1¶

- `country: optional string or null`¶

The two letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2) of the user.¶

maxLength: 2, minLength: 2¶

- `region: optional string or null`¶

The region of the user.¶

maxLength: 255, minLength: 1¶

- `timezone: optional string or null`¶

The [IANA timezone](https://nodatime.org/TimeZones) of the user.¶

maxLength: 255, minLength: 1¶

### Web Search Tool 20260209¶

- `WebSearchTool20260209 object`¶

- `type: "web_search_20260209"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `type: "approximate"`¶

- `city: optional string or null`¶

The city of the user.¶

maxLength: 255, minLength: 1¶

- `country: optional string or null`¶

The two letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2) of the user.¶

maxLength: 2, minLength: 2¶

- `region: optional string or null`¶

The region of the user.¶

maxLength: 255, minLength: 1¶

- `timezone: optional string or null`¶

The [IANA timezone](https://nodatime.org/TimeZones) of the user.¶

maxLength: 255, minLength: 1¶

### Web Search Tool 20260318¶

- `WebSearchTool20260318 object`¶

- `type: "web_search_20260318"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `type: "approximate"`¶

- `city: optional string or null`¶

The city of the user.¶

maxLength: 255, minLength: 1¶

- `country: optional string or null`¶

The two letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2) of the user.¶

maxLength: 2, minLength: 2¶

- `region: optional string or null`¶

The region of the user.¶

maxLength: 255, minLength: 1¶

- `timezone: optional string or null`¶

The [IANA timezone](https://nodatime.org/TimeZones) of the user.¶

maxLength: 255, minLength: 1¶

### Web Search Tool Request Error¶

- `WebSearchToolRequestError object`¶

- `type: "web_search_tool_result_error"`¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

### Web Search Tool Result Block¶

- `WebSearchToolResultBlock object`¶

- `type: "web_search_tool_result"`¶

default: web_search_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `content: WebSearchToolResultBlockContent`¶

- `WebSearchToolResultError object`¶

- `type: "web_search_tool_result_error"`¶

default: web_search_tool_result_error¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `array of WebSearchResultBlock`¶

- `type: "web_search_result"`¶

default: web_search_result¶

- `encrypted_content: string`¶

- `page_age: string or null`¶

- `title: string`¶

- `url: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Web Search Tool Result Block Content¶

- `WebSearchToolResultBlockContent = WebSearchToolResultError or array of WebSearchResultBlock`¶

- `WebSearchToolResultError object`¶

- `type: "web_search_tool_result_error"`¶

default: web_search_tool_result_error¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `array of WebSearchResultBlock`¶

- `type: "web_search_result"`¶

default: web_search_result¶

- `encrypted_content: string`¶

- `page_age: string or null`¶

- `title: string`¶

- `url: string`¶

### Web Search Tool Result Block Param¶

- `WebSearchToolResultBlockParam object`¶

- `type: "web_search_tool_result"`¶

- `content: WebSearchToolResultBlockParamContent`¶

- `WebSearchToolResultBlockItem = array of WebSearchResultBlockParam`¶

- `type: "web_search_result"`¶

- `encrypted_content: string`¶

- `title: string`¶

- `url: string`¶

- `page_age: optional string or null`¶

- `WebSearchToolRequestError object`¶

- `type: "web_search_tool_result_error"`¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

### Web Search Tool Result Block Param Content¶

- `WebSearchToolResultBlockParamContent = array of WebSearchResultBlockParam or WebSearchToolRequestError`¶

- `WebSearchToolResultBlockItem = array of WebSearchResultBlockParam`¶

- `type: "web_search_result"`¶

- `encrypted_content: string`¶

- `title: string`¶

- `url: string`¶

- `page_age: optional string or null`¶

- `WebSearchToolRequestError object`¶

- `type: "web_search_tool_result_error"`¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

### Web Search Tool Result Error¶

- `WebSearchToolResultError object`¶

- `type: "web_search_tool_result_error"`¶

default: web_search_tool_result_error¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

### Web Search Tool Result Error Code¶

- `WebSearchToolResultErrorCode = "invalid_tool_input" or "unavailable" or "max_uses_exceeded" or 3 more`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

## Messages › Batches¶

### Create a Message Batch¶

**POST** `/v1/messages/batches`¶

Send a batch of Message creation requests.¶

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.¶

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing)¶

#### Headers¶

- `"anthropic-user-profile-id": optional string`¶

The user profile ID to attribute the requests in this batch to. Use when acting on behalf of a party other than your organization. Requires the `user-profiles` beta header. Applies to every request in the batch; an individual request whose `user_profile_id` body field conflicts with this header is errored.¶

- `"anthropic-workspace-id": optional string`¶

#### Body parameters¶

- `requests: array of object`¶

List of requests for prompt completion. Each is an individual request to create a Message.¶

maxItems: 100000, minItems: 1¶

- `custom_id: string`¶

Developer-provided ID created for each request in a Message Batch. Useful for matching results to requests, as results may be given out of request order.¶

Must be unique for each request within the Message Batch.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,64}$¶

- `params: object`¶

Messages API creation parameters for the individual request.¶

See the [Messages API reference](https://platform.claude.com/docs/en/api/messages) for full documentation on available parameters.¶

- `max_tokens: number`¶

The maximum number of tokens to generate before stopping.¶

Note that our models may stop _before_ reaching this maximum. This parameter only specifies the absolute maximum number of tokens to generate.¶

Set to `0` to populate the [prompt cache](https://platform.claude.com/docs/en/build-with-claude/prompt-caching#pre-warming-the-cache) without generating a response.¶

Different models have different maximum values for this parameter. See [models](https://platform.claude.com/docs/en/about-claude/models/overview) for details.¶

minimum: 0¶

- `messages: array of MessageParam`¶

Input messages.¶

Our models are trained to operate on alternating `user` and `assistant` conversational turns. When creating a new `Message`, you specify the prior conversational turns with the `messages` parameter, and the model then generates the next `Message` in the conversation. Consecutive `user` or `assistant` turns in your request will be combined into a single turn.¶

Each input message must be an object with a `role` and `content`. You can specify a single `user`-role message, or you can include multiple `user` and `assistant` messages.¶

If the final message uses the `assistant` role, the response content will continue immediately from the content in that message. This can be used to constrain part of the model's response.¶

Example with a single `user` message:¶

```json¶
[{"role": "user", "content": "Hello, Claude"}]¶
```¶

Example with multiple conversational turns:¶

```json¶
[¶
{"role": "user", "content": "Hello there."},¶
{"role": "assistant", "content": "Hi, I'm Claude. How can I help you?"},¶
{"role": "user", "content": "Can you explain LLMs in plain English?"},¶
]¶
```¶

Example with a partially-filled response from Claude:¶

```json¶
[¶
{"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},¶
{"role": "assistant", "content": "The best answer is ("},¶
]¶
```¶

Each input message `content` may be either a single `string` or an array of content blocks, where each block has a specific `type`. Using a `string` for `content` is shorthand for an array of one content block of type `"text"`. The following input messages are equivalent:¶

```json¶
{"role": "user", "content": "Hello, Claude"}¶
```¶

```json¶
{"role": "user", "content": [{"type": "text", "text": "Hello, Claude"}]}¶
```¶

See [input examples](https://platform.claude.com/docs/en/build-with-claude/working-with-messages).¶

Note that if you want to include a [system prompt](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role), you can use the top-level `system` parameter — there is no `"system"` role for input messages in the Messages API.¶

There is a limit of 100,000 messages in a single request.¶

- `content: string or array of ContentBlockParam`¶

- `string`¶

- `array of ContentBlockParam`¶

- `TextBlockParam object`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `type: "ephemeral"`¶

- `ttl: optional "5m" or "1h"`¶

The time-to-live for the cache control breakpoint.¶

This may be one the following values:¶

- `5m`: 5 minutes¶
- `1h`: 1 hour¶

Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching) for details.¶

- `"5m"`¶

- `"1h"`¶

- `citations: optional array of TextCitationParam or null`¶

- `CitationCharLocationParam object`¶

- `type: "char_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_char_index: number`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocationParam object`¶

- `type: "page_location"`¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_page_number: number`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocationParam object`¶

- `type: "content_block_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

maxLength: 500, minLength: 1¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationWebSearchResultLocationParam object`¶

- `type: "web_search_result_location"`¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512, minLength: 1¶

- `url: string`¶

minLength: 1¶

- `CitationSearchResultLocationParam object`¶

- `type: "search_result_location"`¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `ImageBlockParam object`¶

- `type: "image"`¶

- `source: Base64ImageSource or URLImageSource or FileImageSource`¶

- `Base64ImageSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "image/jpeg" or "image/png" or "image/gif" or "image/webp"`¶

- `"image/jpeg"`¶

- `"image/png"`¶

- `"image/gif"`¶

- `"image/webp"`¶

- `URLImageSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileImageSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `transformations: optional ImageTransformationsParam or null`¶

Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.¶

- `oversized_image: optional "downsize" or "error"`¶

What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.¶

- `"downsize"`¶

- `"error"`¶

- `DocumentBlockParam object`¶

- `type: "document"`¶

- `source: Base64PDFSource or PlainTextSource or ContentBlockSource or 2 more`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `ContentBlockSource object`¶

- `type: "content"`¶

- `content: string or array of ContentBlockSourceContent`¶

- `string`¶

- `ContentBlockSourceContent = array of ContentBlockSourceContent`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `URLPDFSource object`¶

- `type: "url"`¶

- `url: string`¶

- `FileDocumentSource object`¶

- `type: "file"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

- `enabled: optional boolean`¶

- `context: optional string or null`¶

minLength: 1¶

- `title: optional string or null`¶

maxLength: 500, minLength: 1¶

- `SearchResultBlockParam object`¶

- `type: "search_result"`¶

- `content: array of TextBlockParam`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional array of TextCitationParam or null`¶

- `source: string`¶

- `title: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam`¶

- `ThinkingBlockParam object`¶

- `type: "thinking"`¶

- `signature: string`¶

The `signature` value of this thinking block, exactly as returned by the API in a previous response. Used to verify that the block was generated by Claude.¶

Thinking blocks must be passed back unmodified and in their original order; a modified block results in a 400 `invalid_request_error`.¶

- `thinking: string`¶

The `thinking` text of this block as returned by the API.¶

- `RedactedThinkingBlockParam object`¶

- `type: "redacted_thinking"`¶

- `data: string`¶

The `data` value of this redacted thinking block, exactly as returned by the API in a previous response. Opaque and encrypted; pass it back unchanged.¶

- `ToolUseBlockParam object`¶

- `type: "tool_use"`¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `input: map[unknown]`¶

- `name: string`¶

maxLength: 200, minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family this member belongs to.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ToolResultBlockParam object`¶

- `type: "tool_result"`¶

- `tool_use_id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `content: optional string or array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `string`¶

- `array of TextBlockParam or ImageBlockParam or SearchResultBlockParam or 3 more`¶

- `TextBlockParam object`¶

- `ImageBlockParam object`¶

- `SearchResultBlockParam object`¶

- `DocumentBlockParam object`¶

- `ToolReferenceBlockParam object`¶

Tool reference block that can be included in tool_result content.¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BrowserStateBlockParam object`¶

The caller's browser state after a browser toolset member call —¶
the full inventory of open tabs, which tab is active, and any side¶
effects (tabs opened, download state changes) the call produced.¶

At most one per `tool_result`, only on a non-error result answering a¶
browser toolset member `tool_use`. The server renders the¶
model-visible text from it; the model never sees the raw fields.¶

- `type: "browser_state"`¶

- `tabs: array of BrowserStateTabEntry`¶

All tabs open in the browser after this call — the full inventory, not a delta. May be empty. Whenever non-empty, exactly one entry carries `active: true`.¶

maxItems: 100¶

- `tab_id: string`¶

The caller-assigned identifier for this tab, unique within the inventory.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `title: string`¶

The title of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The URL of the page the tab is showing. May be empty.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `active: optional boolean`¶

Whether this tab is the active tab after this call. Whenever `tabs` is non-empty, exactly one entry is marked `active: true`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `state_changes: optional array of BrowserStateChange or null`¶

Tabs opened and download state changes during this call. "Nothing to report" is expressed by omitting the field, never by an empty list.¶

maxItems: 200, minItems: 1¶

- `BrowserStateChangeTabOpened object`¶

A tab this call's execution opened that remains open at its end —¶
the creation delta of the `tabs` inventory, not an event log.¶

Carries only the `tab_id`; the tab's `title` and `url` live on its¶
`tabs` entry, which must include the same `tab_id`. A tab opened¶
during a failed call gets no deferred `tab_opened`; it simply appears¶
in the next result's `tabs` inventory.¶

- `type: "tab_opened"`¶

- `tab_id: string`¶

The `tab_id` of the opened tab, present in `tabs`.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadStarted object`¶

A file download that started during this call.¶

- `type: "download_started"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `BrowserStateChangeDownloadCompleted object`¶

A file download that finished during this call, reported with the¶
same `download_id` as its `download_started` — or without a prior¶
`download_started`, when the download finished during the call that¶
started it (at most one state change per `download_id` per result).¶

- `type: "download_completed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `path: optional string or null`¶

Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `size_bytes: optional number or null`¶

The completed download's size.¶

minimum: 0¶

- `BrowserStateChangeDownloadFailed object`¶

A file download that failed — or was cancelled — during this call.¶

- `type: "download_failed"`¶

- `download_id: string`¶

The caller-assigned identifier for this download, stable across the state changes reporting it.¶

maxLength: 4096, minLength: 1, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `url: string`¶

The final post-redirect URL the download was served from.¶

maxLength: 4096, pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$¶

- `error: optional string or null`¶

The failure or cancellation detail, when known.¶

pattern: ^[^\x00-\x1f\x7f-\x9f\u2028\u2029]*$, maxLength: 4096¶

- `is_error: optional boolean`¶

- `toolset_name: optional string or null`¶

For a toolset member tool_result, the toolset family of the paired tool_use.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlockParam object`¶

- `type: "server_tool_use"`¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `WebSearchToolResultBlockParam object`¶

- `type: "web_search_tool_result"`¶

- `content: WebSearchToolResultBlockParamContent`¶

- `WebSearchToolResultBlockItem = array of WebSearchResultBlockParam`¶

- `type: "web_search_result"`¶

- `encrypted_content: string`¶

- `title: string`¶

- `url: string`¶

- `page_age: optional string or null`¶

- `WebSearchToolRequestError object`¶

- `type: "web_search_tool_result_error"`¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `WebFetchToolResultBlockParam object`¶

- `type: "web_fetch_tool_result"`¶

- `content: WebFetchToolResultErrorBlockParam or WebFetchBlockParam`¶

- `WebFetchToolResultErrorBlockParam object`¶

- `type: "web_fetch_tool_result_error"`¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlockParam object`¶

- `type: "web_fetch_result"`¶

- `content: DocumentBlockParam`¶

- `url: string`¶

Fetched content URL¶

- `retrieved_at: optional string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `caller: optional DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `CodeExecutionToolResultBlockParam object`¶

- `type: "code_execution_tool_result"`¶

- `content: CodeExecutionToolResultBlockParamContent`¶

- `CodeExecutionToolResultErrorParam object`¶

- `type: "code_execution_tool_result_error"`¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlockParam object`¶

- `type: "code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlockParam object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

- `content: array of CodeExecutionOutputBlockParam`¶

- `type: "code_execution_output"`¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `BashCodeExecutionToolResultBlockParam object`¶

- `type: "bash_code_execution_tool_result"`¶

- `content: BashCodeExecutionToolResultErrorParam or BashCodeExecutionResultBlockParam`¶

- `BashCodeExecutionToolResultErrorParam object`¶

- `type: "bash_code_execution_tool_result_error"`¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlockParam object`¶

- `type: "bash_code_execution_result"`¶

- `content: array of BashCodeExecutionOutputBlockParam`¶

- `type: "bash_code_execution_output"`¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `TextEditorCodeExecutionToolResultBlockParam object`¶

- `type: "text_editor_code_execution_tool_result"`¶

- `content: TextEditorCodeExecutionToolResultErrorParam or TextEditorCodeExecutionViewResultBlockParam or TextEditorCodeExecutionCreateResultBlockParam or TextEditorCodeExecutionStrReplaceResultBlockParam`¶

- `TextEditorCodeExecutionToolResultErrorParam object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: optional string or null`¶

- `TextEditorCodeExecutionViewResultBlockParam object`¶

- `type: "text_editor_code_execution_view_result"`¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: optional number or null`¶

- `start_line: optional number or null`¶

- `total_lines: optional number or null`¶

- `TextEditorCodeExecutionCreateResultBlockParam object`¶

- `type: "text_editor_code_execution_create_result"`¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlockParam object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

- `lines: optional array of string or null`¶

- `new_lines: optional number or null`¶

- `new_start: optional number or null`¶

- `old_lines: optional number or null`¶

- `old_start: optional number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `ToolSearchToolResultBlockParam object`¶

- `type: "tool_search_tool_result"`¶

- `content: ToolSearchToolResultErrorParam or ToolSearchToolSearchResultBlockParam`¶

- `ToolSearchToolResultErrorParam object`¶

- `type: "tool_search_tool_result_error"`¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: optional string or null`¶

- `ToolSearchToolSearchResultBlockParam object`¶

- `type: "tool_search_tool_search_result"`¶

- `tool_references: array of ToolReferenceBlockParam`¶

- `type: "tool_reference"`¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `ContainerUploadBlockParam object`¶

A content block that represents a file to be uploaded to the container¶
Files uploaded via this block will be available in the container's input directory.¶

- `type: "container_upload"`¶

- `file_id: string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `role: "user" or "assistant" or "system"`¶

- `"user"`¶

- `"assistant"`¶

- `"system"`¶

- `model: Model`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1"`¶

Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows¶

- `"claude-mythos-5-1"`¶

Our most capable model for cybersecurity and biology research, available through trusted access programs¶

- `"claude-sonnet-5"`¶

High-performance model for coding and agents¶

- `"claude-fable-5"`¶

Next generation of intelligence for the hardest knowledge work and coding problems¶

- `"claude-mythos-5"`¶

Most capable model for cybersecurity and biology research¶

- `"claude-opus-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-8"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-7"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-mythos-preview"`¶

New class of intelligence, strongest in coding and cybersecurity¶

- `"claude-opus-4-6"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-6"`¶

Best combination of speed and intelligence¶

- `"claude-haiku-4-5"`¶

Fastest model with near-frontier intelligence¶

- `"claude-haiku-4-5-20251001"`¶

Fastest model with near-frontier intelligence¶

- `"claude-opus-4-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-5-20251101"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-5"`¶

High-performance model for agents and coding¶

- `"claude-sonnet-4-5-20250929"`¶

High-performance model for agents and coding¶

- `string`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Top-level cache control automatically applies a cache_control marker to the last cacheable block in the request.¶

- `container: optional MessageCreateParamsContainer or null`¶

Container identifier for reuse across requests.¶

- `ContainerParams object`¶

Container parameters with skills to be loaded.¶

- `id: optional string or null`¶

Container id¶

- `skills: optional array of SkillParams or null`¶

List of skills to load in the container¶

maxItems: 20¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: optional string`¶

Skill version or 'latest' for most recent version¶

maxLength: 64, minLength: 1¶

- `string`¶

- `inference_geo: optional string or null`¶

Specifies the geographic region for inference processing. If not specified, the workspace's `default_inference_geo` is used.¶

- `metadata: optional Metadata`¶

An object describing metadata about the request.¶

- `user_id: optional string or null`¶

An external identifier for the user who is associated with the request.¶

This should be a uuid, hash value, or other opaque identifier. Anthropic may use this id to help detect abuse. Do not include any identifying information such as name, email address, or phone number.¶

maxLength: 512¶

- `output_config: optional OutputConfig`¶

Configuration options for the model's output, such as the output format.¶

- `effort: optional "low" or "medium" or "high" or 2 more or null`¶

All possible effort levels.¶

- `"low"`¶

- `"medium"`¶

- `"high"`¶

- `"xhigh"`¶

- `"max"`¶

- `format: optional JSONOutputFormat or null`¶

A schema to specify Claude's output format in responses. See [structured outputs](https://platform.claude.com/docs/en/build-with-claude/structured-outputs)¶

- `type: "json_schema"`¶

- `schema: map[unknown]`¶

The JSON schema of the format¶

- `service_tier: optional "auto" or "standard_only"`¶

Determines whether to use priority capacity (if available) or standard capacity for this request.¶

Anthropic offers different levels of service for your API requests. See [service-tiers](https://platform.claude.com/docs/en/api/service-tiers) for details.¶

- `"auto"`¶

- `"standard_only"`¶

- `stop_sequences: optional array of string`¶

Custom text sequences that will cause the model to stop generating.¶

Our models will normally stop when they have naturally completed their turn, which will result in a response `stop_reason` of `"end_turn"`.¶

If you want the model to stop generating when it encounters custom strings of text, you can use the `stop_sequences` parameter. If the model encounters one of the custom sequences, the response `stop_reason` value will be `"stop_sequence"` and the response `stop_sequence` value will contain the matched stop sequence.¶

- `stream: optional boolean`¶

Whether to incrementally stream the response using server-sent events.¶

See [streaming](https://platform.claude.com/docs/en/build-with-claude/streaming) for details.¶

- `system: optional string or array of TextBlockParam`¶

System prompt.¶

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role. See our [guide to system prompts](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices#give-claude-a-role).¶

- `string`¶

- `array of TextBlockParam`¶

- `type: "text"`¶

- `text: string`¶

minLength: 1¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional array of TextCitationParam or null`¶

- `thinking: optional ThinkingConfigParam`¶

Configuration for enabling Claude's extended thinking.¶

When enabled, responses include `thinking` content blocks showing Claude's thinking process before the final answer. Requires a minimum budget of 1,024 tokens and counts towards your `max_tokens` limit.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `ThinkingConfigEnabled object`¶

- `type: "enabled"`¶

- `budget_tokens: number`¶

Determines how many tokens Claude can use for its internal reasoning process. Larger budgets can enable more thorough analysis for complex problems, improving response quality.¶

Must be ≥1024 and less than `max_tokens`.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

minimum: 1024¶

- `display: optional "summarized" or "omitted" or null`¶

Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.¶

- `"summarized"`¶

- `"omitted"`¶

- `ThinkingConfigDisabled object`¶

- `type: "disabled"`¶

- `ThinkingConfigAdaptive object`¶

- `type: "adaptive"`¶

- `display: optional "summarized" or "omitted" or null`¶

Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. Defaults to `summarized`.¶

- `"summarized"`¶

- `"omitted"`¶

- `tool_choice: optional ToolChoice`¶

How the model should use the provided tools. The model can use a specific tool, any available tool, decide by itself, or not use tools at all.¶

- `ToolChoiceAuto object`¶

The model will automatically decide whether to use tools.¶

- `type: "auto"`¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output at most one tool use.¶

- `ToolChoiceAny object`¶

The model will use any available tools.¶

- `type: "any"`¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output exactly one tool use.¶

- `ToolChoiceTool object`¶

The model will use the specified tool with `tool_choice.name`.¶

- `type: "tool"`¶

- `name: string`¶

The name of the tool to use.¶

- `disable_parallel_tool_use: optional boolean`¶

Whether to disable parallel tool use.¶

Defaults to `false`. If set to `true`, the model will output exactly one tool use.¶

- `ToolChoiceNone object`¶

The model will not be allowed to use tools.¶

- `type: "none"`¶

- `tools: optional array of ToolUnion`¶

Definitions of tools that the model may use.¶

If you include `tools` in your API request, the model may return `tool_use` content blocks that represent the model's use of those tools. You can then run those tools using the tool input generated by the model and then optionally return results back to the model using `tool_result` content blocks.¶

There are two types of tools: **client tools** and **server tools**. The behavior described below applies to client tools. For [server tools](https://platform.claude.com/docs/en/agents-and-tools/tool-use/server-tools), see their individual documentation as each has its own behavior (e.g., the [web search tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-search-tool)).¶

Each tool definition includes:¶

* `name`: Name of the tool.¶
* `description`: Optional, but strongly-recommended description of the tool.¶
* `input_schema`: [JSON schema](https://json-schema.org/draft/2020-12) for the tool `input` shape that the model will produce in `tool_use` output content blocks.¶

For example, if you defined `tools` as:¶

```json¶
[¶
{¶
"name": "get_stock_price",¶
"description": "Get the current stock price for a given ticker symbol.",¶
"input_schema": {¶
"type": "object",¶
"properties": {¶
"ticker": {¶
"type": "string",¶
"description": "The stock ticker symbol, e.g. AAPL for Apple Inc."¶
}¶
},¶
"required": ["ticker"]¶
}¶
}¶
]¶
```¶

And then asked the model "What's the S&P 500 at today?", the model might produce `tool_use` content blocks in the response like this:¶

```json¶
[¶
{¶
"type": "tool_use",¶
"id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",¶
"name": "get_stock_price",¶
"input": { "ticker": "^GSPC" }¶
}¶
]¶
```¶

You might then run your `get_stock_price` tool with `{"ticker": "^GSPC"}` as an input, and return the following back to the model in a subsequent `user` message:¶

```json¶
[¶
{¶
"type": "tool_result",¶
"tool_use_id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",¶
"content": "259.75 USD"¶
}¶
]¶
```¶

Tools can be used for workflows that include running client-side tools and functions, or more generally whenever you want the model to produce a particular JSON structure of output.¶

See our [guide](https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview) for more details.¶

- `Tool object`¶

- `type: optional "custom" or null`¶

- `input_schema: object`¶

[JSON schema](https://json-schema.org/draft/2020-12) for this tool's input.¶

This defines the shape of the `input` that your tool accepts and that the model will produce.¶

- `type: "object"`¶

- `properties: optional map[unknown] or null`¶

- `required: optional array of string or null`¶

- `name: string`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

maxLength: 128, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,128}$¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `description: optional string`¶

Description of what this tool does.¶

Tool descriptions should be as detailed as possible. The more information that the model has about what the tool is and how to use it, the better it will perform. You can use natural language descriptions to reinforce important aspects of the tool input JSON schema.¶

- `eager_input_streaming: optional boolean or null`¶

Enable eager input streaming for this tool. When true, tool input parameters will be streamed incrementally as they are generated, and types will be inferred on-the-fly rather than buffering the full JSON output. When false, streaming is disabled for this tool even if the fine-grained-tool-streaming beta is active. When null (default), uses the default behavior based on beta headers.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolBash20250124 object`¶

- `type: "bash_20250124"`¶

- `name: "bash"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20250522 object`¶

- `type: "code_execution_20250522"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20250825 object`¶

- `type: "code_execution_20250825"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20260120 object`¶

Code execution tool with REPL state persistence (daemon mode + gVisor checkpoint).¶

- `type: "code_execution_20260120"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `CodeExecutionTool20260521 object`¶

Code execution tool with REPL state persistence.¶

- `type: "code_execution_20260521"`¶

- `name: "code_execution"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `BrowserToolset20260801 object`¶

The browser toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the browser tool family. The model is served¶
the family's tool with any members disabled via `configs` removed¶
from its schema.¶

- `type: "browser_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `configs: optional BrowserToolsetConfigs or null`¶

Per-member configuration for `browser_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional BrowserTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `close_tab: optional BrowserCloseTabConfig or null`¶

`close_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional BrowserDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `file_upload: optional BrowserFileUploadConfig or null`¶

`file_upload`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `find: optional BrowserFindConfig or null`¶

`find`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `form_input: optional BrowserFormInputConfig or null`¶

`form_input`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `get_page_text: optional BrowserGetPageTextConfig or null`¶

`get_page_text`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional BrowserHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hover: optional BrowserHoverConfig or null`¶

`hover`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `javascript_exec: optional BrowserJavascriptExecConfig or null`¶

`javascript_exec`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional BrowserKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional BrowserLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional BrowserLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional BrowserLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional BrowserLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `list_tabs: optional BrowserListTabsConfig or null`¶

`list_tabs`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional BrowserMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional BrowserMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `navigate: optional BrowserNavigateConfig or null`¶

`navigate`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `new_tab: optional BrowserNewTabConfig or null`¶

`new_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_console: optional BrowserReadConsoleConfig or null`¶

`read_console`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_network: optional BrowserReadNetworkConfig or null`¶

`read_network`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `read_page: optional BrowserReadPageConfig or null`¶

`read_page`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional BrowserRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional BrowserScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional BrowserScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll_to: optional BrowserScrollToConfig or null`¶

`scroll_to`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `switch_tab: optional BrowserSwitchTabConfig or null`¶

`switch_tab`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional BrowserTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional BrowserWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional BrowserZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `MemoryTool20250818 object`¶

- `type: "memory_20250818"`¶

- `name: "memory"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ComputerToolset20260801 object`¶

The computer toolset: a single `tools[]` entry (carrying no¶
`name`) that declares the computer tool family. The model is¶
served the family's tool with any members disabled via `configs`¶
removed from its schema. Every member is enabled by default, zoom¶
included. The single-tool options `display_number` and¶
`enable_zoom` are not fields of a toolset entry — it carries only¶
`type`, `configs`, and `cache_control`; zoom is controlled¶
via `configs.zoom.enabled`.¶

- `type: "computer_toolset_20260801"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `configs: optional ComputerToolsetConfigs or null`¶

Per-member configuration for `computer_toolset_20260801`: one¶
optional field per member tool, keyed by the member name — the same¶
name the member's `tool_use` blocks carry. Every member is an¶
accepted key, and a member's defaults apply wherever its key is¶
absent. Unknown keys are rejected: the field set is this toolset¶
version's complete member set.¶

- `type: optional ComputerTypeConfig or null`¶

`type`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `cursor_position: optional ComputerCursorPositionConfig or null`¶

`cursor_position`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `double_click: optional ComputerDoubleClickConfig or null`¶

`double_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `hold_key: optional ComputerHoldKeyConfig or null`¶

`hold_key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `key: optional ComputerKeyConfig or null`¶

`key`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click: optional ComputerLeftClickConfig or null`¶

`left_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_click_drag: optional ComputerLeftClickDragConfig or null`¶

`left_click_drag`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_down: optional ComputerLeftMouseDownConfig or null`¶

`left_mouse_down`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `left_mouse_up: optional ComputerLeftMouseUpConfig or null`¶

`left_mouse_up`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `middle_click: optional ComputerMiddleClickConfig or null`¶

`middle_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `mouse_move: optional ComputerMouseMoveConfig or null`¶

`mouse_move`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `right_click: optional ComputerRightClickConfig or null`¶

`right_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `screenshot: optional ComputerScreenshotConfig or null`¶

`screenshot`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `scroll: optional ComputerScrollConfig or null`¶

`scroll`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `triple_click: optional ComputerTripleClickConfig or null`¶

`triple_click`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `wait: optional ComputerWaitConfig or null`¶

`wait`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `zoom: optional ComputerZoomConfig or null`¶

`zoom`'s config overrides.¶

- `defer_loading: optional boolean or null`¶

Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.¶

- `enabled: optional boolean or null`¶

Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.¶

- `ToolTextEditor20250124 object`¶

- `type: "text_editor_20250124"`¶

- `name: "str_replace_editor"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolTextEditor20250429 object`¶

- `type: "text_editor_20250429"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolTextEditor20250728 object`¶

- `type: "text_editor_20250728"`¶

- `name: "str_replace_based_edit_tool"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `input_examples: optional array of map[unknown]`¶

- `max_characters: optional number or null`¶

Maximum number of characters to display when viewing a file. If not specified, defaults to displaying the full file.¶

minimum: 1¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `WebSearchTool20250305 object`¶

- `type: "web_search_20250305"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `type: "approximate"`¶

- `city: optional string or null`¶

The city of the user.¶

maxLength: 255, minLength: 1¶

- `country: optional string or null`¶

The two letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2) of the user.¶

maxLength: 2, minLength: 2¶

- `region: optional string or null`¶

The region of the user.¶

maxLength: 255, minLength: 1¶

- `timezone: optional string or null`¶

The [IANA timezone](https://nodatime.org/TimeZones) of the user.¶

maxLength: 255, minLength: 1¶

- `WebFetchTool20250910 object`¶

- `type: "web_fetch_20250910"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

- `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `type: "all"`¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `type: "none"`¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `type: "only"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `type: "except"`¶

- `tools: array of WebFetchURLSourceToolReference`¶

- `type: "tool_reference"`¶

- `name: string`¶

- `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`¶

Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

- `WebFetchURLSourceOnly object`¶

The tool filter variant under which only the named tools' results¶
contribute.¶

- `WebFetchURLSourceExcept object`¶

The tool filter variant under which every result but the named¶
tools' contributes.¶

- `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`¶

Whether URLs in user messages are fetchable: "all" or "none".¶

- `WebFetchURLSourceAll object`¶

The `url_sources` variant under which a source contributes in¶
full: every result of the tool filter's source, or all user input.¶

- `WebFetchURLSourceNone object`¶

The `url_sources` variant under which a source contributes nothing:¶
no result of the tool filter's source, or no user input.¶

 - `WebSearchTool20260209 object`¶

- `type: "web_search_20260209"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `WebFetchTool20260209 object`¶

- `type: "web_fetch_20260209"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `WebFetchTool20260309 object`¶

Web fetch tool with use_cache parameter for bypassing cached content.¶

- `type: "web_fetch_20260309"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

- `WebSearchTool20260318 object`¶

- `type: "web_search_20260318"`¶

- `name: "web_search"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.¶

- `blocked_domains: optional array of string or null`¶

If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `user_location: optional UserLocation or null`¶

Parameters for the user's location. Used to provide more relevant search results.¶

- `WebFetchTool20260318 object`¶

- `type: "web_fetch_20260318"`¶

- `name: "web_fetch"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `allowed_domains: optional array of string or null`¶

List of domains to allow fetching from¶

- `blocked_domains: optional array of string or null`¶

List of domains to block fetching from¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `citations: optional CitationsConfigParam or null`¶

Citations configuration for fetched documents. Citations are disabled by default.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `max_content_tokens: optional number or null`¶

Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.¶

exclusiveMinimum: 0¶

- `max_uses: optional number or null`¶

Maximum number of times the tool can be used in the API request.¶

exclusiveMinimum: 0¶

- `response_inclusion: optional "full" or "excluded"`¶

How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.¶

- `"full"`¶

- `"excluded"`¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

 - `url_sources: optional WebFetchURLSources or null`¶

Which sources contribute to the set of URLs web fetch may fetch.¶

Each key is a tagged variant: `user_input` is `all` or `none`; the¶
two tool filters are `all`, `none`, `only` (only the named tools'¶
results) or `except` (every result but the named tools'). A named tool¶
must be declared in this request's `tools[]`.¶

 - `use_cache: optional boolean`¶

Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.¶

- `ToolSearchToolBm25_20251119 object`¶

- `type: "tool_search_tool_bm25_20251119" or "tool_search_tool_bm25"`¶

- `"tool_search_tool_bm25_20251119"`¶

- `"tool_search_tool_bm25"`¶

- `name: "tool_search_tool_bm25"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `ToolSearchToolRegex20251119 object`¶

- `type: "tool_search_tool_regex_20251119" or "tool_search_tool_regex"`¶

- `"tool_search_tool_regex_20251119"`¶

- `"tool_search_tool_regex"`¶

- `name: "tool_search_tool_regex"`¶

Name of the tool.¶

This is how the tool will be called by the model and in `tool_use` blocks.¶

- `allowed_callers: optional array of "direct" or "code_execution_20250825" or "code_execution_20260120" or "code_execution_20260521"`¶

- `"direct"`¶

- `"code_execution_20250825"`¶

- `"code_execution_20260120"`¶

- `"code_execution_20260521"`¶

- `cache_control: optional CacheControlEphemeral or null`¶

Create a cache control breakpoint at this content block.¶

- `defer_loading: optional boolean`¶

If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.¶

- `strict: optional boolean`¶

When true, guarantees schema validation on tool names and inputs¶

- `temperature: optional number`¶

**Deprecated**: Deprecated. Models released after Claude Opus 4.6 do not support setting temperature. A value of 1.0 will be accepted for backwards compatibility, all other values will be rejected with a 400 error.¶

Amount of randomness injected into the response.¶

Defaults to `1.0`. Ranges from `0.0` to `1.0`. Use `temperature` closer to `0.0` for analytical / multiple choice, and closer to `1.0` for creative and generative tasks.¶

Note that even with `temperature` of `0.0`, the results will not be fully deterministic.¶

maximum: 1, minimum: 0¶

- `top_k: optional number`¶

**Deprecated**: Deprecated. Models released after Claude Opus 4.6 do not accept top_k; any value will be rejected with a 400 error.¶

Only sample from the top K options for each subsequent token.¶

Used to remove "long tail" low probability responses. [Learn more technical details here](https://towardsdatascience.com/how-to-sample-from-language-models-682bceb97277).¶

Recommended for advanced use cases only.¶

minimum: 0¶

- `top_p: optional number`¶

**Deprecated**: Deprecated. Models released after Claude Opus 4.6 do not support setting top_p. A value >= 0.99 will be accepted for backwards compatibility, all other values will be rejected with a 400 error.¶

Use nucleus sampling.¶

In nucleus sampling, we compute the cumulative distribution over all the options for each subsequent token in decreasing probability order and cut it off once it reaches a particular probability specified by `top_p`.¶

Recommended for advanced use cases only.¶

maximum: 1, minimum: 0¶

#### Returns¶

- `MessageBatch object`¶

- `type: "message_batch"`¶

Object type.¶

For Message Batches, this is always `"message_batch"`.¶

default: message_batch¶

- `id: string`¶

Unique object identifier.¶

The format and length of IDs may change over time.¶

- `archived_at: string or null`¶

RFC 3339 datetime string representing the time at which the Message Batch was archived and its results became unavailable.¶

format: date-time¶

- `cancel_initiated_at: string or null`¶

RFC 3339 datetime string representing the time at which cancellation was initiated for the Message Batch. Specified only if cancellation was initiated.¶

format: date-time¶

- `created_at: string`¶

RFC 3339 datetime string representing the time at which the Message Batch was created.¶

format: date-time¶

- `ended_at: string or null`¶

RFC 3339 datetime string representing the time at which processing for the Message Batch ended. Specified only once processing ends.¶

Processing ends when every request in a Message Batch has either succeeded, errored, canceled, or expired.¶

format: date-time¶

- `expires_at: string`¶

RFC 3339 datetime string representing the time at which the Message Batch will expire and end processing, which is 24 hours after creation.¶

format: date-time¶

- `processing_status: "in_progress" or "canceling" or "ended"`¶

Processing status of the Message Batch.¶

- `"in_progress"`¶

- `"canceling"`¶

- `"ended"`¶

- `request_counts: MessageBatchRequestCounts`¶

Tallies requests within the Message Batch, categorized by their status.¶

Requests start as `processing` and move to one of the other statuses only once processing of the entire batch ends. The sum of all values always matches the total number of requests in the batch.¶

- `canceled: number`¶

Number of requests in the Message Batch that have been canceled.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `errored: number`¶

Number of requests in the Message Batch that encountered an error.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `expired: number`¶

Number of requests in the Message Batch that have expired.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `processing: number`¶

Number of requests in the Message Batch that are processing.¶

default: 0¶

- `succeeded: number`¶

Number of requests in the Message Batch that have completed successfully.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `results_url: string or null`¶

URL to a `.jsonl` file containing the results of the Message Batch requests. Specified only once processing ends.¶

Results in the file are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.¶

#### Example¶

```bash¶
curl https://api.anthropic.com/v1/messages/batches \¶
-H 'Content-Type: application/json' \¶
-H 'anthropic-version: 2023-06-01' \¶
-H "X-Api-Key: $ANTHROPIC_API_KEY" \¶
-d '{¶
"requests": [¶
{¶
"custom_id": "my-custom-id-1",¶
"params": {¶
"max_tokens": 1024,¶
"messages": [¶
{¶
"content": "Hello, world",¶
"role": "user"¶
}¶
],¶
"model": "claude-opus-5"¶
}¶
}¶
]¶
}'¶
```¶

##### Response (200)¶

```json¶
{¶
"id": "msgbatch_013Zva2CMHLNnXjNJJKqJ2EF",¶
"archived_at": "2024-08-20T18:37:24.100435Z",¶
"cancel_initiated_at": "2024-08-20T18:37:24.100435Z",¶
"created_at": "2024-08-20T18:37:24.100435Z",¶
"ended_at": "2024-08-20T18:37:24.100435Z",¶
"expires_at": "2024-08-20T18:37:24.100435Z",¶
"processing_status": "in_progress",¶
"request_counts": {¶
"canceled": 10,¶
"errored": 30,¶
"expired": 10,¶
"processing": 100,¶
"succeeded": 50¶
},¶
"results_url": "https://api.anthropic.com/v1/messages/batches/msgbatch_013Zva2CMHLNnXjNJJKqJ2EF/results",¶
"type": "message_batch"¶
}¶
```¶

### Retrieve a Message Batch¶

**GET** `/v1/messages/batches/{message_batch_id}`¶

This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.¶

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing)¶

#### Path parameters¶

- `message_batch_id: string`¶

ID of the Message Batch.¶

#### Headers¶

- `"anthropic-workspace-id": optional string`¶

#### Returns¶

- `MessageBatch object`¶

- `type: "message_batch"`¶

Object type.¶

For Message Batches, this is always `"message_batch"`.¶

default: message_batch¶

- `id: string`¶

Unique object identifier.¶

The format and length of IDs may change over time.¶

- `archived_at: string or null`¶

RFC 3339 datetime string representing the time at which the Message Batch was archived and its results became unavailable.¶

format: date-time¶

- `cancel_initiated_at: string or null`¶

RFC 3339 datetime string representing the time at which cancellation was initiated for the Message Batch. Specified only if cancellation was initiated.¶

format: date-time¶

- `created_at: string`¶

RFC 3339 datetime string representing the time at which the Message Batch was created.¶

format: date-time¶

- `ended_at: string or null`¶

RFC 3339 datetime string representing the time at which processing for the Message Batch ended. Specified only once processing ends.¶

Processing ends when every request in a Message Batch has either succeeded, errored, canceled, or expired.¶

format: date-time¶

- `expires_at: string`¶

RFC 3339 datetime string representing the time at which the Message Batch will expire and end processing, which is 24 hours after creation.¶

format: date-time¶

- `processing_status: "in_progress" or "canceling" or "ended"`¶

Processing status of the Message Batch.¶

- `"in_progress"`¶

- `"canceling"`¶

- `"ended"`¶

- `request_counts: MessageBatchRequestCounts`¶

Tallies requests within the Message Batch, categorized by their status.¶

Requests start as `processing` and move to one of the other statuses only once processing of the entire batch ends. The sum of all values always matches the total number of requests in the batch.¶

- `canceled: number`¶

Number of requests in the Message Batch that have been canceled.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `errored: number`¶

Number of requests in the Message Batch that encountered an error.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `expired: number`¶

Number of requests in the Message Batch that have expired.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `processing: number`¶

Number of requests in the Message Batch that are processing.¶

default: 0¶

- `succeeded: number`¶

Number of requests in the Message Batch that have completed successfully.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `results_url: string or null`¶

URL to a `.jsonl` file containing the results of the Message Batch requests. Specified only once processing ends.¶

Results in the file are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.¶

#### Example¶

```bash¶
curl https://api.anthropic.com/v1/messages/batches/$MESSAGE_BATCH_ID \¶
-H 'anthropic-version: 2023-06-01' \¶
-H "X-Api-Key: $ANTHROPIC_API_KEY"¶
```¶

##### Response (200)¶

```json¶
{¶
"id": "msgbatch_013Zva2CMHLNnXjNJJKqJ2EF",¶
"archived_at": "2024-08-20T18:37:24.100435Z",¶
"cancel_initiated_at": "2024-08-20T18:37:24.100435Z",¶
"created_at": "2024-08-20T18:37:24.100435Z",¶
"ended_at": "2024-08-20T18:37:24.100435Z",¶
"expires_at": "2024-08-20T18:37:24.100435Z",¶
"processing_status": "in_progress",¶
"request_counts": {¶
"canceled": 10,¶
"errored": 30,¶
"expired": 10,¶
"processing": 100,¶
"succeeded": 50¶
},¶
"results_url": "https://api.anthropic.com/v1/messages/batches/msgbatch_013Zva2CMHLNnXjNJJKqJ2EF/results",¶
"type": "message_batch"¶
}¶
```¶

### List Message Batches¶

**GET** `/v1/messages/batches`¶

List all Message Batches within a Workspace. Most recently created batches are returned first.¶

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing)¶

#### Query parameters¶

- `after_id: optional string`¶

ID of the object to use as a cursor for pagination. When provided, returns the page of results immediately after this object.¶

- `before_id: optional string`¶

ID of the object to use as a cursor for pagination. When provided, returns the page of results immediately before this object.¶

- `limit: optional number`¶

Number of items to return per page.¶

Defaults to `20`. Ranges from `1` to `1000`.¶

default: 20, maximum: 1000, minimum: 1¶

#### Headers¶

- `"anthropic-workspace-id": optional string`¶

#### Returns¶

- `data: array of MessageBatch`¶

- `type: "message_batch"`¶

Object type.¶

For Message Batches, this is always `"message_batch"`.¶

default: message_batch¶

- `id: string`¶

Unique object identifier.¶

The format and length of IDs may change over time.¶

- `archived_at: string or null`¶

RFC 3339 datetime string representing the time at which the Message Batch was archived and its results became unavailable.¶

format: date-time¶

- `cancel_initiated_at: string or null`¶

RFC 3339 datetime string representing the time at which cancellation was initiated for the Message Batch. Specified only if cancellation was initiated.¶

format: date-time¶

- `created_at: string`¶

RFC 3339 datetime string representing the time at which the Message Batch was created.¶

format: date-time¶

- `ended_at: string or null`¶

RFC 3339 datetime string representing the time at which processing for the Message Batch ended. Specified only once processing ends.¶

Processing ends when every request in a Message Batch has either succeeded, errored, canceled, or expired.¶

format: date-time¶

- `expires_at: string`¶

RFC 3339 datetime string representing the time at which the Message Batch will expire and end processing, which is 24 hours after creation.¶

format: date-time¶

- `processing_status: "in_progress" or "canceling" or "ended"`¶

Processing status of the Message Batch.¶

- `"in_progress"`¶

- `"canceling"`¶

- `"ended"`¶

- `request_counts: MessageBatchRequestCounts`¶

Tallies requests within the Message Batch, categorized by their status.¶

Requests start as `processing` and move to one of the other statuses only once processing of the entire batch ends. The sum of all values always matches the total number of requests in the batch.¶

- `canceled: number`¶

Number of requests in the Message Batch that have been canceled.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `errored: number`¶

Number of requests in the Message Batch that encountered an error.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `expired: number`¶

Number of requests in the Message Batch that have expired.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `processing: number`¶

Number of requests in the Message Batch that are processing.¶

default: 0¶

- `succeeded: number`¶

Number of requests in the Message Batch that have completed successfully.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `results_url: string or null`¶

URL to a `.jsonl` file containing the results of the Message Batch requests. Specified only once processing ends.¶

Results in the file are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.¶

- `first_id: string or null`¶

First ID in the `data` list. Can be used as the `before_id` for the previous page.¶

- `has_more: boolean`¶

Indicates if there are more results in the requested page direction.¶

- `last_id: string or null`¶

Last ID in the `data` list. Can be used as the `after_id` for the next page.¶

#### Example¶

```bash¶
curl https://api.anthropic.com/v1/messages/batches \¶
-H 'anthropic-version: 2023-06-01' \¶
-H "X-Api-Key: $ANTHROPIC_API_KEY"¶
```¶

##### Response (200)¶

```json¶
{¶
"data": [¶
{¶
"id": "msgbatch_013Zva2CMHLNnXjNJJKqJ2EF",¶
"archived_at": "2024-08-20T18:37:24.100435Z",¶
"cancel_initiated_at": "2024-08-20T18:37:24.100435Z",¶
"created_at": "2024-08-20T18:37:24.100435Z",¶
"ended_at": "2024-08-20T18:37:24.100435Z",¶
"expires_at": "2024-08-20T18:37:24.100435Z",¶
"processing_status": "in_progress",¶
"request_counts": {¶
"canceled": 10,¶
"errored": 30,¶
"expired": 10,¶
"processing": 100,¶
"succeeded": 50¶
},¶
"results_url": "https://api.anthropic.com/v1/messages/batches/msgbatch_013Zva2CMHLNnXjNJJKqJ2EF/results",¶
"type": "message_batch"¶
}¶
],¶
"first_id": "first_id",¶
"has_more": true,¶
"last_id": "last_id"¶
}¶
```¶

### Cancel a Message Batch¶

**POST** `/v1/messages/batches/{message_batch_id}/cancel`¶

Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.¶

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.¶

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing)¶

#### Path parameters¶

- `message_batch_id: string`¶

ID of the Message Batch.¶

#### Headers¶

- `"anthropic-workspace-id": optional string`¶

#### Returns¶

- `MessageBatch object`¶

- `type: "message_batch"`¶

Object type.¶

For Message Batches, this is always `"message_batch"`.¶

default: message_batch¶

- `id: string`¶

Unique object identifier.¶

The format and length of IDs may change over time.¶

- `archived_at: string or null`¶

RFC 3339 datetime string representing the time at which the Message Batch was archived and its results became unavailable.¶

format: date-time¶

- `cancel_initiated_at: string or null`¶

RFC 3339 datetime string representing the time at which cancellation was initiated for the Message Batch. Specified only if cancellation was initiated.¶

format: date-time¶

- `created_at: string`¶

RFC 3339 datetime string representing the time at which the Message Batch was created.¶

format: date-time¶

- `ended_at: string or null`¶

RFC 3339 datetime string representing the time at which processing for the Message Batch ended. Specified only once processing ends.¶

Processing ends when every request in a Message Batch has either succeeded, errored, canceled, or expired.¶

format: date-time¶

- `expires_at: string`¶

RFC 3339 datetime string representing the time at which the Message Batch will expire and end processing, which is 24 hours after creation.¶

format: date-time¶

- `processing_status: "in_progress" or "canceling" or "ended"`¶

Processing status of the Message Batch.¶

- `"in_progress"`¶

- `"canceling"`¶

- `"ended"`¶

- `request_counts: MessageBatchRequestCounts`¶

Tallies requests within the Message Batch, categorized by their status.¶

Requests start as `processing` and move to one of the other statuses only once processing of the entire batch ends. The sum of all values always matches the total number of requests in the batch.¶

- `canceled: number`¶

Number of requests in the Message Batch that have been canceled.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `errored: number`¶

Number of requests in the Message Batch that encountered an error.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `expired: number`¶

Number of requests in the Message Batch that have expired.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `processing: number`¶

Number of requests in the Message Batch that are processing.¶

default: 0¶

- `succeeded: number`¶

Number of requests in the Message Batch that have completed successfully.¶

This is zero until processing of the entire Message Batch has ended.¶

default: 0¶

- `results_url: string or null`¶

URL to a `.jsonl` file containing the results of the Message Batch requests. Specified only once processing ends.¶

Results in the file are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.¶

#### Example¶

```bash¶
curl https://api.anthropic.com/v1/messages/batches/$MESSAGE_BATCH_ID/cancel \¶
-X POST \¶
-H 'anthropic-version: 2023-06-01' \¶
-H "X-Api-Key: $ANTHROPIC_API_KEY"¶
```¶

##### Response (200)¶

```json¶
{¶
"id": "msgbatch_013Zva2CMHLNnXjNJJKqJ2EF",¶
"archived_at": "2024-08-20T18:37:24.100435Z",¶
"cancel_initiated_at": "2024-08-20T18:37:24.100435Z",¶
"created_at": "2024-08-20T18:37:24.100435Z",¶
"ended_at": "2024-08-20T18:37:24.100435Z",¶
"expires_at": "2024-08-20T18:37:24.100435Z",¶
"processing_status": "in_progress",¶
"request_counts": {¶
"canceled": 10,¶
"errored": 30,¶
"expired": 10,¶
"processing": 100,¶
"succeeded": 50¶
},¶
"results_url": "https://api.anthropic.com/v1/messages/batches/msgbatch_013Zva2CMHLNnXjNJJKqJ2EF/results",¶
"type": "message_batch"¶
}¶
```¶

### Delete a Message Batch¶

**DELETE** `/v1/messages/batches/{message_batch_id}`¶

Delete a Message Batch.¶

Message Batches can only be deleted once they've finished processing. If you'd like to delete an in-progress batch, you must first cancel it.¶

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing)¶

#### Path parameters¶

- `message_batch_id: string`¶

ID of the Message Batch.¶

#### Headers¶

- `"anthropic-workspace-id": optional string`¶

#### Returns¶

- `DeletedMessageBatch object`¶

- `type: "message_batch_deleted"`¶

Deleted object type.¶

For Message Batches, this is always `"message_batch_deleted"`.¶

default: message_batch_deleted¶

- `id: string`¶

ID of the Message Batch.¶

#### Example¶

```bash¶
curl https://api.anthropic.com/v1/messages/batches/$MESSAGE_BATCH_ID \¶
-X DELETE \¶
-H 'anthropic-version: 2023-06-01' \¶
-H "X-Api-Key: $ANTHROPIC_API_KEY"¶
```¶

##### Response (200)¶

```json¶
{¶
"id": "msgbatch_013Zva2CMHLNnXjNJJKqJ2EF",¶
"type": "message_batch_deleted"¶
}¶
```¶

### Retrieve Message Batch results¶

**GET** `/v1/messages/batches/{message_batch_id}/results`¶

Streams the results of a Message Batch as a `.jsonl` file.¶

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.¶

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing)¶

#### Path parameters¶

- `message_batch_id: string`¶

ID of the Message Batch.¶

#### Headers¶

- `"anthropic-workspace-id": optional string`¶

#### Returns¶

- `MessageBatchIndividualResponse object`¶

This is a single line in the response `.jsonl` file and does not represent the response as a whole.¶

- `custom_id: string`¶

Developer-provided ID created for each request in a Message Batch. Useful for matching results to requests, as results may be given out of request order.¶

Must be unique for each request within the Message Batch.¶

- `result: MessageBatchResult`¶

Processing result for this request.¶

Contains a Message output if processing was successful, an error response if processing failed, or the reason why processing was not attempted, such as cancellation or expiration.¶

- `MessageBatchSucceededResult object`¶

- `type: "succeeded"`¶

default: succeeded¶

- `message: Message`¶

- `type: "message"`¶

Object type.¶

For Messages, this is always `"message"`.¶

default: message¶

- `id: string`¶

Unique object identifier.¶

The format and length of IDs may change over time.¶

- `container: Container or null`¶

Information about the container used in the request (for the code execution tool)¶

- `id: string`¶

Identifier for the container used in this request¶

- `expires_at: string`¶

The time at which the container will expire.¶

format: date-time¶

- `skills: array of ContainerSkill or null`¶

Skills loaded in the container¶

- `type: "anthropic" or "custom"`¶

Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)¶

- `"anthropic"`¶

- `"custom"`¶

- `skill_id: string`¶

Skill ID¶

maxLength: 64, minLength: 1¶

- `version: string`¶

The resolved version: a skill version ID for custom skills.¶

maxLength: 64, minLength: 1¶

- `content: array of ContentBlock`¶

Content generated by the model.¶

This is an array of content blocks, each of which has a `type` that determines its shape.¶

Example:¶

```json¶
[{"type": "text", "text": "Hi, I'm Claude."}]¶
```¶

If the request input `messages` ended with an `assistant` turn, then the response `content` will continue directly from that last turn. You can use this to constrain the model's output.¶

For example, if the input `messages` were:¶

```json¶
[¶
{"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},¶
{"role": "assistant", "content": "The best answer is ("}¶
]¶
```¶

Then the response `content` might be:¶

```json¶
[{"type": "text", "text": "B)"}]¶
```¶

- `TextBlock object`¶

- `type: "text"`¶

default: text¶

- `citations: array of TextCitation or null`¶

Citations supporting the text block.¶

The type of citation returned will depend on the type of document being cited. Citing a PDF results in `page_location`, plain text results in `char_location`, and content document results in `content_block_location`.¶

- `CitationCharLocation object`¶

- `type: "char_location"`¶

default: char_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_char_index: number`¶

- `file_id: string or null`¶

- `start_char_index: number`¶

minimum: 0¶

- `CitationPageLocation object`¶

- `type: "page_location"`¶

default: page_location¶

- `cited_text: string`¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_page_number: number`¶

- `file_id: string or null`¶

- `start_page_number: number`¶

minimum: 1¶

- `CitationContentBlockLocation object`¶

- `type: "content_block_location"`¶

default: content_block_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `document_index: number`¶

minimum: 0¶

- `document_title: string or null`¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `file_id: string or null`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `CitationsWebSearchResultLocation object`¶

- `type: "web_search_result_location"`¶

default: web_search_result_location¶

- `cited_text: string`¶

- `encrypted_index: string`¶

- `title: string or null`¶

maxLength: 512¶

- `url: string`¶

- `CitationsSearchResultLocation object`¶

- `type: "search_result_location"`¶

default: search_result_location¶

- `cited_text: string`¶

The full text of the cited block range, concatenated.¶

Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.¶

- `end_block_index: number`¶

Exclusive 0-based end index of the cited block range in the source's `content` array.¶

Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.¶

- `search_result_index: number`¶

0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.¶

Counted separately from `document_index`; server-side web search results are not included in this count.¶

minimum: 0¶

- `source: string`¶

- `start_block_index: number`¶

0-based index of the first cited block in the source's `content` array.¶

minimum: 0¶

- `title: string or null`¶

- `text: string`¶

minLength: 0¶

- `ThinkingBlock object`¶

- `type: "thinking"`¶

default: thinking¶

- `signature: string`¶

A value used to verify that this thinking block was generated by Claude when it is passed back to the API.¶

This is an opaque field and should not be interpreted or parsed. When passing thinking blocks back to the API (required when using tools with extended thinking), pass them back exactly as received, with this field intact.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking) for details.¶

- `thinking: string`¶

The text of Claude's thinking process for this block.¶

- `RedactedThinkingBlock object`¶

- `type: "redacted_thinking"`¶

default: redacted_thinking¶

- `data: string`¶

The contents of this redacted thinking block, returned when portions of the model's thinking were safety-redacted. This field is opaque and encrypted, with no readable content.¶

Pass `redacted_thinking` blocks back to the API unchanged when continuing a multi-turn conversation.¶

See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking#redacted-thinking-blocks) for details.¶

- `ToolUseBlock object`¶

- `type: "tool_use"`¶

default: tool_use¶

- `id: string`¶

pattern: ^[a-zA-Z0-9_-]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `type: "direct"`¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `type: "code_execution_20250825"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ServerToolCaller20260120 object`¶

- `type: "code_execution_20260120"`¶

- `tool_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `input: map[unknown]`¶

- `name: string`¶

minLength: 1¶

- `toolset_name: optional string or null`¶

For a toolset member tool_use, the toolset family.¶

maxLength: 64, minLength: 1, pattern: ^[a-zA-Z0-9_-]+$¶

- `ServerToolUseBlock object`¶

- `type: "server_tool_use"`¶

default: server_tool_use¶

- `id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `input: map[unknown]`¶

- `name: "web_search" or "web_fetch" or "code_execution" or 4 more`¶

- `"web_search"`¶

- `"web_fetch"`¶

- `"code_execution"`¶

- `"bash_code_execution"`¶

- `"text_editor_code_execution"`¶

- `"tool_search_tool_regex"`¶

- `"tool_search_tool_bm25"`¶

- `WebSearchToolResultBlock object`¶

- `type: "web_search_tool_result"`¶

default: web_search_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebSearchToolResultBlockContent`¶

- `WebSearchToolResultError object`¶

- `type: "web_search_tool_result_error"`¶

default: web_search_tool_result_error¶

- `error_code: WebSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"max_uses_exceeded"`¶

- `"too_many_requests"`¶

- `"query_too_long"`¶

- `"request_too_large"`¶

- `array of WebSearchResultBlock`¶

- `type: "web_search_result"`¶

default: web_search_result¶

- `encrypted_content: string`¶

- `page_age: string or null`¶

- `title: string`¶

- `url: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `WebFetchToolResultBlock object`¶

- `type: "web_fetch_tool_result"`¶

default: web_fetch_tool_result¶

- `caller: DirectCaller or ServerToolCaller or ServerToolCaller20260120`¶

default: {"type":"direct"}¶

- `DirectCaller object`¶

Tool invocation directly from the model.¶

- `ServerToolCaller object`¶

Tool invocation generated by a server-side tool.¶

- `ServerToolCaller20260120 object`¶

- `content: WebFetchToolResultErrorBlock or WebFetchBlock`¶

- `WebFetchToolResultErrorBlock object`¶

- `type: "web_fetch_tool_result_error"`¶

default: web_fetch_tool_result_error¶

- `error_code: WebFetchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"url_too_long"`¶

- `"url_not_allowed"`¶

- `"url_not_in_prior_context"`¶

- `"url_not_accessible"`¶

- `"unsupported_content_type"`¶

- `"too_many_requests"`¶

- `"max_uses_exceeded"`¶

- `"unavailable"`¶

- `"content_too_large"`¶

- `WebFetchBlock object`¶

- `type: "web_fetch_result"`¶

default: web_fetch_result¶

- `content: DocumentBlock`¶

- `type: "document"`¶

default: document¶

- `citations: CitationsConfig or null`¶

Citation configuration for the document¶

- `enabled: boolean`¶

default: false¶

- `source: Base64PDFSource or PlainTextSource`¶

- `Base64PDFSource object`¶

- `type: "base64"`¶

- `data: string`¶

format: byte¶

- `media_type: "application/pdf"`¶

- `PlainTextSource object`¶

- `type: "text"`¶

- `data: string`¶

- `media_type: "text/plain"`¶

- `title: string or null`¶

The title of the document¶

- `retrieved_at: string or null`¶

ISO 8601 timestamp when the content was retrieved¶

- `url: string`¶

Fetched content URL¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `CodeExecutionToolResultBlock object`¶

- `type: "code_execution_tool_result"`¶

default: code_execution_tool_result¶

- `content: CodeExecutionToolResultBlockContent`¶

- `CodeExecutionToolResultError object`¶

- `type: "code_execution_tool_result_error"`¶

default: code_execution_tool_result_error¶

- `error_code: CodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `CodeExecutionResultBlock object`¶

- `type: "code_execution_result"`¶

default: code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `EncryptedCodeExecutionResultBlock object`¶

Code execution result with encrypted stdout for PFC + web_search results.¶

- `type: "encrypted_code_execution_result"`¶

default: encrypted_code_execution_result¶

- `content: array of CodeExecutionOutputBlock`¶

- `type: "code_execution_output"`¶

default: code_execution_output¶

- `file_id: string`¶

- `encrypted_stdout: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `BashCodeExecutionToolResultBlock object`¶

- `type: "bash_code_execution_tool_result"`¶

default: bash_code_execution_tool_result¶

- `content: BashCodeExecutionToolResultError or BashCodeExecutionResultBlock`¶

- `BashCodeExecutionToolResultError object`¶

- `type: "bash_code_execution_tool_result_error"`¶

default: bash_code_execution_tool_result_error¶

- `error_code: BashCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"output_file_too_large"`¶

- `BashCodeExecutionResultBlock object`¶

- `type: "bash_code_execution_result"`¶

default: bash_code_execution_result¶

- `content: array of BashCodeExecutionOutputBlock`¶

- `type: "bash_code_execution_output"`¶

default: bash_code_execution_output¶

- `file_id: string`¶

- `return_code: number`¶

- `stderr: string`¶

- `stdout: string`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `TextEditorCodeExecutionToolResultBlock object`¶

- `type: "text_editor_code_execution_tool_result"`¶

default: text_editor_code_execution_tool_result¶

- `content: TextEditorCodeExecutionToolResultError or TextEditorCodeExecutionViewResultBlock or TextEditorCodeExecutionCreateResultBlock or TextEditorCodeExecutionStrReplaceResultBlock`¶

- `TextEditorCodeExecutionToolResultError object`¶

- `type: "text_editor_code_execution_tool_result_error"`¶

default: text_editor_code_execution_tool_result_error¶

- `error_code: TextEditorCodeExecutionToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `"file_not_found"`¶

- `error_message: string or null`¶

- `TextEditorCodeExecutionViewResultBlock object`¶

- `type: "text_editor_code_execution_view_result"`¶

default: text_editor_code_execution_view_result¶

- `content: string`¶

- `file_type: "text" or "image" or "pdf"`¶

- `"text"`¶

- `"image"`¶

- `"pdf"`¶

- `num_lines: number or null`¶

- `start_line: number or null`¶

- `total_lines: number or null`¶

- `TextEditorCodeExecutionCreateResultBlock object`¶

- `type: "text_editor_code_execution_create_result"`¶

default: text_editor_code_execution_create_result¶

- `is_file_update: boolean`¶

- `TextEditorCodeExecutionStrReplaceResultBlock object`¶

- `type: "text_editor_code_execution_str_replace_result"`¶

default: text_editor_code_execution_str_replace_result¶

- `lines: array of string or null`¶

- `new_lines: number or null`¶

- `new_start: number or null`¶

- `old_lines: number or null`¶

- `old_start: number or null`¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ToolSearchToolResultBlock object`¶

- `type: "tool_search_tool_result"`¶

default: tool_search_tool_result¶

- `content: ToolSearchToolResultError or ToolSearchToolSearchResultBlock`¶

- `ToolSearchToolResultError object`¶

- `type: "tool_search_tool_result_error"`¶

default: tool_search_tool_result_error¶

- `error_code: ToolSearchToolResultErrorCode`¶

- `"invalid_tool_input"`¶

- `"unavailable"`¶

- `"too_many_requests"`¶

- `"execution_time_exceeded"`¶

- `error_message: string or null`¶

- `ToolSearchToolSearchResultBlock object`¶

- `type: "tool_search_tool_search_result"`¶

default: tool_search_tool_search_result¶

- `tool_references: array of ToolReferenceBlock`¶

- `type: "tool_reference"`¶

default: tool_reference¶

- `tool_name: string`¶

maxLength: 256, minLength: 1, pattern: ^[a-zA-Z0-9_-]{1,256}$¶

- `tool_use_id: string`¶

pattern: ^srvtoolu_[a-zA-Z0-9_]+$¶

- `ContainerUploadBlock object`¶

Response model for a file uploaded to the container.¶

- `type: "container_upload"`¶

default: container_upload¶

- `file_id: string`¶

- `model: Model`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more`¶

The model that will complete your prompt.¶

See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.¶

- `"claude-fable-5-1"`¶

Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows¶

- `"claude-mythos-5-1"`¶

Our most capable model for cybersecurity and biology research, available through trusted access programs¶

- `"claude-sonnet-5"`¶

High-performance model for coding and agents¶

- `"claude-fable-5"`¶

Next generation of intelligence for the hardest knowledge work and coding problems¶

- `"claude-mythos-5"`¶

Most capable model for cybersecurity and biology research¶

- `"claude-opus-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-8"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-7"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-mythos-preview"`¶

New class of intelligence, strongest in coding and cybersecurity¶

- `"claude-opus-4-6"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-6"`¶

Best combination of speed and intelligence¶

- `"claude-haiku-4-5"`¶

Fastest model with near-frontier intelligence¶

- `"claude-haiku-4-5-20251001"`¶

Fastest model with near-frontier intelligence¶

- `"claude-opus-4-5"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-opus-4-5-20251101"`¶

Powerful intelligence for long-running agents and coding¶

- `"claude-sonnet-4-5"`¶

High-performance model for agents and coding¶

- `"claude-sonnet-4-5-20250929"`¶

High-performance model for agents and coding¶

- `string`¶

- `role: "assistant"`¶

Conversational role of the generated message.¶

This will always be `"assistant"`.¶

default: assistant¶

- `stop_details: RefusalStopDetails or null`¶

Structured information about a refusal.¶

- `type: "refusal"`¶

default: refusal¶

- `category: "cyber" or "bio" or "frontier_llm" or 2 more or null`¶

The policy category that triggered a refusal.¶

- `"cyber"`¶

The request could enable cyber harm, such as malware or exploit development. Benign cybersecurity work can also trigger this category.¶

- `"bio"`¶

The request could enable biological harm, such as dangerous lab methods. Beneficial life sciences work can also trigger this category.¶

- `"frontier_llm"`¶

The request could assist the development of competing AI models, which is restricted under [Anthropic's commercial terms](https://www.anthropic.com/legal/commercial-terms). Benign machine learning work can also trigger this category.¶

- `"reasoning_extraction"`¶

The request asks the model to reproduce its internal reasoning in the response text. To get reasoning in a structured form instead, use [adaptive thinking](https://platform.claude.com/docs/en/build-with-claude/adaptive-thinking).¶

- `"general_harms"`¶

The request could be related to an area that was determined as harmful. Benign work might sometimes trigger this category.¶

- `explanation: string or null`¶

Human-readable explanation of the refusal.¶

This text is not guaranteed to be stable. `null` when no explanation is available for the category.¶

- `stop_reason: StopReason or null`¶

The reason that we stopped.¶

This may be one the following values:¶

* `"end_turn"`: the model reached a natural stopping point¶
* `"max_tokens"`: we exceeded the requested `max_tokens` or the model's maximum¶
* `"stop_sequence"`: one of your provided custom `stop_sequences` was generated¶
* `"tool_use"`: the model invoked one or more tools¶
* `"pause_turn"`: we paused a long-running turn. You may provide the response back as-is in a subsequent request to let the model continue.¶
* `"refusal"`: when streaming classifiers intervene to handle potential policy violations¶
* `"model_context_window_exceeded"`: we exceeded the model's context window¶

In non-streaming mode this value is always non-null. In streaming mode, it is null in the `message_start` event and non-null otherwise.¶

- `"end_turn"`¶

- `"max_tokens"`¶

- `"stop_sequence"`¶

- `"tool_use"`¶

- `"pause_turn"`¶

- `"refusal"`¶

- `"model_context_window_exceeded"`¶

- `stop_sequence: string or null`¶

Which custom stop sequence was generated, if any.¶

This value will be a non-null string if one of your custom stop sequences was generated.¶

- `usage: Usage`¶

Billing and rate-limit usage.¶

Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.¶

Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.¶

For example, `output_tokens` will be non-zero, even for an empty string response from Claude.¶

Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.¶

- `cache_creation: CacheCreation or null`¶

Breakdown of cached tokens by TTL¶

- `ephemeral_1h_input_tokens: number`¶

The number of input tokens used to create the 1 hour cache entry.¶

default: 0, minimum: 0¶

- `ephemeral_5m_input_tokens: number`¶

The number of input tokens used to create the 5 minute cache entry.¶

default: 0, minimum: 0¶

- `cache_creation_input_tokens: number or null`¶

The number of input tokens used to create the cache entry.¶

minimum: 0¶

- `cache_read_input_tokens: number or null`¶

The number of input tokens read from the cache.¶

minimum: 0¶

- `inference_geo: string or null`¶

The geographic region where inference was performed for this request.¶

- `input_tokens: number`¶

The number of input tokens which were used.¶

minimum: 0¶

- `output_tokens: number`¶

The number of output tokens which were used.¶

minimum: 0¶

- `output_tokens_details: OutputTokensDetails or null`¶

Breakdown of output tokens by category.¶

`output_tokens` remains the inclusive, authoritative total used for billing.¶
This object provides a read-only decomposition for observability — for example,¶
how many of the billed output tokens were spent on internal reasoning that may¶
have been summarized before being returned to you.¶

- `thinking_tokens: number`¶

Number of output tokens the model generated as internal reasoning, including¶
the thinking-block delimiter tokens.¶

Reflects the raw reasoning the model produced, not the (possibly shorter)¶
summarized thinking text returned in the response body. Computed by¶
re-tokenizing the raw reasoning text, so it may differ from the model's exact¶
generation count by a small number of tokens. Always ≤ `output_tokens`;¶
`output_tokens - thinking_tokens` approximates the non-reasoning output.¶

default: 0, minimum: 0¶

- `server_tool_use: ServerToolUsage or null`¶

The number of server tool requests.¶

- `web_fetch_requests: number`¶

The number of web fetch tool requests.¶

default: 0, minimum: 0¶

- `web_search_requests: number`¶

The number of web search tool requests.¶

default: 0, minimum: 0¶

- `service_tier: "standard" or "priority" or "batch" or null`¶

If the request used the priority, standard, or batch tier.¶

- `"standard"`¶

- `"priority"`¶

- `"batch"`¶

- `MessageBatchErroredResult object`¶

- `type: "errored"`¶

default: errored¶

- `error: ErrorResponse`¶

- `type: "error"`¶

default: error¶

- `error: ErrorObject`¶

- `InvalidRequestError object`¶

- `type: "invalid_request_error"`¶

default: invalid_request_error¶

- `message: string`¶

default: Invalid request¶

- `AuthenticationError object`¶

- `type: "authentication_error"`¶

default: authentication_error¶

- `message: string`¶

default: Authentication error¶

- `BillingError object`¶

- `type: "billing_error"`¶

default: billing_error¶

- `message: string`¶

default: Billing error¶

- `PermissionError object`¶

- `type: "permission_error"`¶

default: permission_error¶

- `message: string`¶

default: Permission denied¶

- `NotFoundError object`¶

- `type: "not_found_error"`¶

default: not_found_error¶

- `message: string`¶

default: Not found¶

- `RateLimitError object`¶

- `type: "rate_limit_error"`¶

default: rate_limit_error¶

- `message: string`¶

default: Rate limited¶

- `GatewayTimeoutError object`¶

- `type: "timeout_error"`¶

default: timeout_error¶

- `message: string`¶

default: Request timeout¶

- `APIErrorObject object`¶

- `type: "api_error"`¶

default: api_error¶

- `message: string`¶

default: Internal server error¶

- `OverloadedError object`¶

- `type: "overloaded_error"`¶

default: overloaded_error¶

- `message: string`¶

default: Overloaded¶

- `request_id: string or null`¶

- `MessageBatchCanceledResult object`¶

- `type: "canceled"`¶

default: canceled¶

- `MessageBatchExpiredResult object`¶

- `type: "expired"`¶

default: expired¶

#### Example¶

```bash¶
curl https://api.anthropic.com/v1/messages/batches/$MESSAGE_BATCH_ID/results \¶
-H 'anthropic-version: 2023-06-01' \¶
-H "X-Api-Key: $ANTHROPIC_API_KEY"¶
```¶

Unified Diff

--- a/api/messages.md
+++ b/api/messages.md
@@ -2563,6 +2563,97 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
+      - `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+        Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+          - `type: "all"`
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
+          - `type: "none"`
+
+        - `WebFetchURLSourceOnly object`
+
+          The tool filter variant under which only the named tools' results
+          contribute.
+
+          - `type: "only"`
+
+          - `tools: array of WebFetchURLSourceToolReference`
+
+            - `type: "tool_reference"`
+
+            - `name: string`
+
+        - `WebFetchURLSourceExcept object`
+
+          The tool filter variant under which every result but the named
+          tools' contributes.
+
+          - `type: "except"`
+
+          - `tools: array of WebFetchURLSourceToolReference`
+
+            - `type: "tool_reference"`
+
+            - `name: string`
+
+      - `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+        Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
+        - `WebFetchURLSourceOnly object`
+
+          The tool filter variant under which only the named tools' results
+          contribute.
+
+        - `WebFetchURLSourceExcept object`
+
+          The tool filter variant under which every result but the named
+          tools' contributes.
+
+      - `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`
+
+        Whether URLs in user messages are fetchable: "all" or "none".
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
   - `WebSearchTool20260209 object`
 
     - `type: "web_search_20260209"`
@@ -2669,6 +2760,15 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
   - `WebFetchTool20260309 object`
 
     Web fetch tool with use_cache parameter for bypassing cached content.
@@ -2727,6 +2827,15 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
     - `use_cache: optional boolean`
 
       Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.
@@ -2852,6 +2961,15 @@
     - `strict: optional boolean`
 
       When true, guarantees schema validation on tool names and inputs
+
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
 
     - `use_cache: optional boolean`
 
@@ -6776,6 +6894,97 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
+      - `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+        Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+          - `type: "all"`
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
+          - `type: "none"`
+
+        - `WebFetchURLSourceOnly object`
+
+          The tool filter variant under which only the named tools' results
+          contribute.
+
+          - `type: "only"`
+
+          - `tools: array of WebFetchURLSourceToolReference`
+
+            - `type: "tool_reference"`
+
+            - `name: string`
+
+        - `WebFetchURLSourceExcept object`
+
+          The tool filter variant under which every result but the named
+          tools' contributes.
+
+          - `type: "except"`
+
+          - `tools: array of WebFetchURLSourceToolReference`
+
+            - `type: "tool_reference"`
+
+            - `name: string`
+
+      - `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+        Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
+        - `WebFetchURLSourceOnly object`
+
+          The tool filter variant under which only the named tools' results
+          contribute.
+
+        - `WebFetchURLSourceExcept object`
+
+          The tool filter variant under which every result but the named
+          tools' contributes.
+
+      - `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`
+
+        Whether URLs in user messages are fetchable: "all" or "none".
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
   - `WebSearchTool20260209 object`
 
     - `type: "web_search_20260209"`
@@ -6882,6 +7091,15 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
   - `WebFetchTool20260309 object`
 
     Web fetch tool with use_cache parameter for bypassing cached content.
@@ -6940,6 +7158,15 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
     - `use_cache: optional boolean`
 
       Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.
@@ -7065,6 +7292,15 @@
     - `strict: optional boolean`
 
       When true, guarantees schema validation on tool names and inputs
+
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
 
     - `use_cache: optional boolean`
 
@@ -15770,6 +16006,97 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
+      - `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+        Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+          - `type: "all"`
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
+          - `type: "none"`
+
+        - `WebFetchURLSourceOnly object`
+
+          The tool filter variant under which only the named tools' results
+          contribute.
+
+          - `type: "only"`
+
+          - `tools: array of WebFetchURLSourceToolReference`
+
+            - `type: "tool_reference"`
+
+            - `name: string`
+
+        - `WebFetchURLSourceExcept object`
+
+          The tool filter variant under which every result but the named
+          tools' contributes.
+
+          - `type: "except"`
+
+          - `tools: array of WebFetchURLSourceToolReference`
+
+            - `type: "tool_reference"`
+
+            - `name: string`
+
+      - `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+        Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
+        - `WebFetchURLSourceOnly object`
+
+          The tool filter variant under which only the named tools' results
+          contribute.
+
+        - `WebFetchURLSourceExcept object`
+
+          The tool filter variant under which every result but the named
+          tools' contributes.
+
+      - `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`
+
+        Whether URLs in user messages are fetchable: "all" or "none".
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
   - `WebSearchTool20260209 object`
 
     - `type: "web_search_20260209"`
@@ -15876,6 +16203,15 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
   - `WebFetchTool20260309 object`
 
     Web fetch tool with use_cache parameter for bypassing cached content.
@@ -15934,6 +16270,15 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
     - `use_cache: optional boolean`
 
       Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.
@@ -16059,6 +16404,15 @@
     - `strict: optional boolean`
 
       When true, guarantees schema validation on tool names and inputs
+
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
 
     - `use_cache: optional boolean`
 
@@ -24804,6 +25158,97 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
+      - `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+        Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+          - `type: "all"`
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
+          - `type: "none"`
+
+        - `WebFetchURLSourceOnly object`
+
+          The tool filter variant under which only the named tools' results
+          contribute.
+
+          - `type: "only"`
+
+          - `tools: array of WebFetchURLSourceToolReference`
+
+            - `type: "tool_reference"`
+
+            - `name: string`
+
+        - `WebFetchURLSourceExcept object`
+
+          The tool filter variant under which every result but the named
+          tools' contributes.
+
+          - `type: "except"`
+
+          - `tools: array of WebFetchURLSourceToolReference`
+
+            - `type: "tool_reference"`
+
+            - `name: string`
+
+      - `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+        Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
+        - `WebFetchURLSourceOnly object`
+
+          The tool filter variant under which only the named tools' results
+          contribute.
+
+        - `WebFetchURLSourceExcept object`
+
+          The tool filter variant under which every result but the named
+          tools' contributes.
+
+      - `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`
+
+        Whether URLs in user messages are fetchable: "all" or "none".
+
+        - `WebFetchURLSourceAll object`
+
+          The `url_sources` variant under which a source contributes in
+          full: every result of the tool filter's source, or all user input.
+
+        - `WebFetchURLSourceNone object`
+
+          The `url_sources` variant under which a source contributes nothing:
+          no result of the tool filter's source, or no user input.
+
   - `WebSearchTool20260209 object`
 
     - `type: "web_search_20260209"`
@@ -24910,6 +25355,15 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
   - `WebFetchTool20260309 object`
 
     Web fetch tool with use_cache parameter for bypassing cached content.
@@ -24968,6 +25422,15 @@
 
       When true, guarantees schema validation on tool names and inputs
 
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
+
     - `use_cache: optional boolean`
 
       Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.
@@ -25093,6 +25556,15 @@
     - `strict: optional boolean`
 
       When true, guarantees schema validation on tool names and inputs
+
+    - `url_sources: optional WebFetchURLSources or null`
+
+      Which sources contribute to the set of URLs web fetch may fetch.
+
+      Each key is a tagged variant: `user_input` is `all` or `none`; the
+      two tool filters are `all`, `none`, `only` (only the named tools'
+      results) or `except` (every result but the named tools'). A named tool
+      must be declared in this request's `tools[]`.
 
     - `use_cache: optional boolean`
 
@@ -25843,6 +26315,97 @@
 
     When true, guarantees schema validation on tool names and inputs
 
+  - `url_sources: optional WebFetchURLSources or null`
+
+    Which sources contribute to the set of URLs web fetch may fetch.
+
+    Each key is a tagged variant: `user_input` is `all` or `none`; the
+    two tool filters are `all`, `none`, `only` (only the named tools'
+    results) or `except` (every result but the named tools'). A named tool
+    must be declared in this request's `tools[]`.
+
+    - `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+      Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+        - `type: "all"`
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
+        - `type: "none"`
+
+      - `WebFetchURLSourceOnly object`
+
+        The tool filter variant under which only the named tools' results
+        contribute.
+
+        - `type: "only"`
+
+        - `tools: array of WebFetchURLSourceToolReference`
+
+          - `type: "tool_reference"`
+
+          - `name: string`
+
+      - `WebFetchURLSourceExcept object`
+
+        The tool filter variant under which every result but the named
+        tools' contributes.
+
+        - `type: "except"`
+
+        - `tools: array of WebFetchURLSourceToolReference`
+
+          - `type: "tool_reference"`
+
+          - `name: string`
+
+    - `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+      Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
+      - `WebFetchURLSourceOnly object`
+
+        The tool filter variant under which only the named tools' results
+        contribute.
+
+      - `WebFetchURLSourceExcept object`
+
+        The tool filter variant under which every result but the named
+        tools' contributes.
+
+    - `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`
+
+      Whether URLs in user messages are fetchable: "all" or "none".
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
 ### Web Fetch Tool 20260209
 
 - `WebFetchTool20260209 object`
@@ -25920,6 +26483,97 @@
 
     When true, guarantees schema validation on tool names and inputs
 
+  - `url_sources: optional WebFetchURLSources or null`
+
+    Which sources contribute to the set of URLs web fetch may fetch.
+
+    Each key is a tagged variant: `user_input` is `all` or `none`; the
+    two tool filters are `all`, `none`, `only` (only the named tools'
+    results) or `except` (every result but the named tools'). A named tool
+    must be declared in this request's `tools[]`.
+
+    - `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+      Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+        - `type: "all"`
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
+        - `type: "none"`
+
+      - `WebFetchURLSourceOnly object`
+
+        The tool filter variant under which only the named tools' results
+        contribute.
+
+        - `type: "only"`
+
+        - `tools: array of WebFetchURLSourceToolReference`
+
+          - `type: "tool_reference"`
+
+          - `name: string`
+
+      - `WebFetchURLSourceExcept object`
+
+        The tool filter variant under which every result but the named
+        tools' contributes.
+
+        - `type: "except"`
+
+        - `tools: array of WebFetchURLSourceToolReference`
+
+          - `type: "tool_reference"`
+
+          - `name: string`
+
+    - `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+      Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
+      - `WebFetchURLSourceOnly object`
+
+        The tool filter variant under which only the named tools' results
+        contribute.
+
+      - `WebFetchURLSourceExcept object`
+
+        The tool filter variant under which every result but the named
+        tools' contributes.
+
+    - `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`
+
+      Whether URLs in user messages are fetchable: "all" or "none".
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
 ### Web Fetch Tool 20260309
 
 - `WebFetchTool20260309 object`
@@ -25999,6 +26653,97 @@
 
     When true, guarantees schema validation on tool names and inputs
 
+  - `url_sources: optional WebFetchURLSources or null`
+
+    Which sources contribute to the set of URLs web fetch may fetch.
+
+    Each key is a tagged variant: `user_input` is `all` or `none`; the
+    two tool filters are `all`, `none`, `only` (only the named tools'
+    results) or `except` (every result but the named tools'). A named tool
+    must be declared in this request's `tools[]`.
+
+    - `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+      Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+        - `type: "all"`
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
+        - `type: "none"`
+
+      - `WebFetchURLSourceOnly object`
+
+        The tool filter variant under which only the named tools' results
+        contribute.
+
+        - `type: "only"`
+
+        - `tools: array of WebFetchURLSourceToolReference`
+
+          - `type: "tool_reference"`
+
+          - `name: string`
+
+      - `WebFetchURLSourceExcept object`
+
+        The tool filter variant under which every result but the named
+        tools' contributes.
+
+        - `type: "except"`
+
+        - `tools: array of WebFetchURLSourceToolReference`
+
+          - `type: "tool_reference"`
+
+          - `name: string`
+
+    - `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+      Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
+      - `WebFetchURLSourceOnly object`
+
+        The tool filter variant under which only the named tools' results
+        contribute.
+
+      - `WebFetchURLSourceExcept object`
+
+        The tool filter variant under which every result but the named
+        tools' contributes.
+
+    - `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`
+
+      Whether URLs in user messages are fetchable: "all" or "none".
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
   - `use_cache: optional boolean`
 
     Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.
@@ -26088,6 +26833,97 @@
 
     When true, guarantees schema validation on tool names and inputs
 
+  - `url_sources: optional WebFetchURLSources or null`
+
+    Which sources contribute to the set of URLs web fetch may fetch.
+
+    Each key is a tagged variant: `user_input` is `all` or `none`; the
+    two tool filters are `all`, `none`, `only` (only the named tools'
+    results) or `except` (every result but the named tools'). A named tool
+    must be declared in this request's `tools[]`.
+
+    - `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+      Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+        - `type: "all"`
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
+        - `type: "none"`
+
+      - `WebFetchURLSourceOnly object`
+
+        The tool filter variant under which only the named tools' results
+        contribute.
+
+        - `type: "only"`
+
+        - `tools: array of WebFetchURLSourceToolReference`
+
+          - `type: "tool_reference"`
+
+          - `name: string`
+
+      - `WebFetchURLSourceExcept object`
+
+        The tool filter variant under which every result but the named
+        tools' contributes.
+
+        - `type: "except"`
+
+        - `tools: array of WebFetchURLSourceToolReference`
+
+          - `type: "tool_reference"`
+
+          - `name: string`
+
+    - `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+      Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
+      - `WebFetchURLSourceOnly object`
+
+        The tool filter variant under which only the named tools' results
+        contribute.
+
+      - `WebFetchURLSourceExcept object`
+
+        The tool filter variant under which every result but the named
+        tools' contributes.
+
+    - `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`
+
+      Whether URLs in user messages are fetchable: "all" or "none".
+
+      - `WebFetchURLSourceAll object`
+
+        The `url_sources` variant under which a source contributes in
+        full: every result of the tool filter's source, or all user input.
+
+      - `WebFetchURLSourceNone object`
+
+        The `url_sources` variant under which a source contributes nothing:
+        no result of the tool filter's source, or no user input.
+
   - `use_cache: optional boolean`
 
     Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.
@@ -26640,6 +27476,158 @@
   - `"unavailable"`
 
   - `"content_too_large"`
+
+### Web Fetch URL Source All
+
+- `WebFetchURLSourceAll object`
+
+  The `url_sources` variant under which a source contributes in
+  full: every result of the tool filter's source, or all user input.
+
+  - `type: "all"`
+
+### Web Fetch URL Source Except
+
+- `WebFetchURLSourceExcept object`
+
+  The tool filter variant under which every result but the named
+  tools' contributes.
+
+  - `type: "except"`
+
+  - `tools: array of WebFetchURLSourceToolReference`
+
+    - `type: "tool_reference"`
+
+    - `name: string`
+
+### Web Fetch URL Source None
+
+- `WebFetchURLSourceNone object`
+
+  The `url_sources` variant under which a source contributes nothing:
+  no result of the tool filter's source, or no user input.
+
+  - `type: "none"`
+
+### Web Fetch URL Source Only
+
+- `WebFetchURLSourceOnly object`
+
+  The tool filter variant under which only the named tools' results
+  contribute.
+
+  - `type: "only"`
+
+  - `tools: array of WebFetchURLSourceToolReference`
+
+    - `type: "tool_reference"`
+
+    - `name: string`
+
+### Web Fetch URL Source Tool Reference
+
+- `WebFetchURLSourceToolReference object`
+
+  One entry of a tool filter's `tools`: it must name a tool declared
+  in this request's `tools[]`.
+
+  - `type: "tool_reference"`
+
+  - `name: string`
+
+### Web Fetch URL Sources
+
+- `WebFetchURLSources object`
+
+  Which sources contribute to the set of URLs web fetch may fetch.
+
+  Each key is a tagged variant: `user_input` is `all` or `none`; the
+  two tool filters are `all`, `none`, `only` (only the named tools'
+  results) or `except` (every result but the named tools'). A named tool
+  must be declared in this request's `tools[]`.
+
+  - `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+    Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].
+
+    - `WebFetchURLSourceAll object`
+
+      The `url_sources` variant under which a source contributes in
+      full: every result of the tool filter's source, or all user input.
+
+      - `type: "all"`
+
+    - `WebFetchURLSourceNone object`
+
+      The `url_sources` variant under which a source contributes nothing:
+      no result of the tool filter's source, or no user input.
+
+      - `type: "none"`
+
+    - `WebFetchURLSourceOnly object`
+
+      The tool filter variant under which only the named tools' results
+      contribute.
+
+      - `type: "only"`
+
+      - `tools: array of WebFetchURLSourceToolReference`
+
+        - `type: "tool_reference"`
+
+        - `name: string`
+
+    - `WebFetchURLSourceExcept object`
+
+      The tool filter variant under which every result but the named
+      tools' contributes.
+
+      - `type: "except"`
+
+      - `tools: array of WebFetchURLSourceToolReference`
+
+        - `type: "tool_reference"`
+
+        - `name: string`
+
+  - `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+    Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.
+
+    - `WebFetchURLSourceAll object`
+
+      The `url_sources` variant under which a source contributes in
+      full: every result of the tool filter's source, or all user input.
+
+    - `WebFetchURLSourceNone object`
+
+      The `url_sources` variant under which a source contributes nothing:
+      no result of the tool filter's source, or no user input.
+
+    - `WebFetchURLSourceOnly object`
+
+      The tool filter variant under which only the named tools' results
+      contribute.
+
+    - `WebFetchURLSourceExcept object`
+
+      The tool filter variant under which every result but the named
+      tools' contributes.
+
+  - `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`
+
+    Whether URLs in user messages are fetchable: "all" or "none".
+
+    - `WebFetchURLSourceAll object`
+
+      The `url_sources` variant under which a source contributes in
+      full: every result of the tool filter's source, or all user input.
+
+    - `WebFetchURLSourceNone object`
+
+      The `url_sources` variant under which a source contributes nothing:
+      no result of the tool filter's source, or no user input.
 
 ### Web Search Result Block
 
@@ -29839,6 +30827,97 @@
 
           When true, guarantees schema validation on tool names and inputs
 
+        - `url_sources: optional WebFetchURLSources or null`
+
+          Which sources contribute to the set of URLs web fetch may fetch.
+
+          Each key is a tagged variant: `user_input` is `all` or `none`; the
+          two tool filters are `all`, `none`, `only` (only the named tools'
+          results) or `except` (every result but the named tools'). A named tool
+          must be declared in this request's `tools[]`.
+
+          - `client_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+            Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].
+
+            - `WebFetchURLSourceAll object`
+
+              The `url_sources` variant under which a source contributes in
+              full: every result of the tool filter's source, or all user input.
+
+              - `type: "all"`
+
+            - `WebFetchURLSourceNone object`
+
+              The `url_sources` variant under which a source contributes nothing:
+              no result of the tool filter's source, or no user input.
+
+              - `type: "none"`
+
+            - `WebFetchURLSourceOnly object`
+
+              The tool filter variant under which only the named tools' results
+              contribute.
+
+              - `type: "only"`
+
+              - `tools: array of WebFetchURLSourceToolReference`
+
+                - `type: "tool_reference"`
+
+                - `name: string`
+
+            - `WebFetchURLSourceExcept object`
+
+              The tool filter variant under which every result but the named
+              tools' contributes.
+
+              - `type: "except"`
+
+              - `tools: array of WebFetchURLSourceToolReference`
+
+                - `type: "tool_reference"`
+
+                - `name: string`
+
+          - `server_tool_results: optional WebFetchURLSourceAll or WebFetchURLSourceNone or WebFetchURLSourceOnly or WebFetchURLSourceExcept`
+
+            Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.
+
+            - `WebFetchURLSourceAll object`
+
+              The `url_sources` variant under which a source contributes in
+              full: every result of the tool filter's source, or all user input.
+
+            - `WebFetchURLSourceNone object`
+
+              The `url_sources` variant under which a source contributes nothing:
+              no result of the tool filter's source, or no user input.
+
+            - `WebFetchURLSourceOnly object`
+
+              The tool filter variant under which only the named tools' results
+              contribute.
+
+            - `WebFetchURLSourceExcept object`
+
+              The tool filter variant under which every result but the named
+              tools' contributes.
+
+          - `user_input: optional WebFetchURLSourceAll or WebFetchURLSourceNone`
+
+            Whether URLs in user messages are fetchable: "all" or "none".
+
+            - `WebFetchURLSourceAll object`
+
+              The `url_sources` variant under which a source contributes in
+              full: every result of the tool filter's source, or all user input.
+
+            - `WebFetchURLSourceNone object`
+
+              The `url_sources` variant under which a source contributes nothing:
+              no result of the tool filter's source, or no user input.
+
       - `WebSearchTool20260209 object`
 
         - `type: "web_search_20260209"`
@@ -29945,6 +31024,15 @@
 
           When true, guarantees schema validation on tool names and inputs
 
+        - `url_sources: optional WebFetchURLSources or null`
+
+          Which sources contribute to the set of URLs web fetch may fetch.
+
+          Each key is a tagged variant: `user_input` is `all` or `none`; the
+          two tool filters are `all`, `none`, `only` (only the named tools'
+          results) or `except` (every result but the named tools'). A named tool
+          must be declared in this request's `tools[]`.
+
       - `WebFetchTool20260309 object`
 
         Web fetch tool with use_cache parameter for bypassing cached content.
@@ -30003,6 +31091,15 @@
 
           When true, guarantees schema validation on tool names and inputs
 
+        - `url_sources: optional WebFetchURLSources or null`
+
+          Which sources contribute to the set of URLs web fetch may fetch.
+
+          Each key is a tagged variant: `user_input` is `all` or `none`; the
+          two tool filters are `all`, `none`, `only` (only the named tools'
+          results) or `except` (every result but the named tools'). A named tool
+          must be declared in this request's `tools[]`.
+
         - `use_cache: optional boolean`
 
           Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.
@@ -30128,6 +31225,15 @@
         - `strict: optional boolean`
 
           When true, guarantees schema validation on tool names and inputs
+
+        - `url_sources: optional WebFetchURLSources or null`
+
+          Which sources contribute to the set of URLs web fetch may fetch.
+
+          Each key is a tagged variant: `user_input` is `all` or `none`; the
+          two tool filters are `all`, `none`, `only` (only the named tools'
+          results) or `except` (every result but the named tools'). A named tool
+          must be declared in this request's `tools[]`.
 
         - `use_cache: optional boolean`