← Back to daily report

build-with-claude/working-with-messages.md

Changed on 2026-02-25 10:57:41 EST

+10 lines added
-17 lines removed
Visual Diff
# Using the Messages API¶

Practical patterns and examples for using the Messages API effectively¶

---¶

This guide covers common patterns for working with the Messages API, including basic requests, multi-turn conversations, prefill techniques, and vision capabilities. For complete API specifications, see the [Messages API reference](/docs/en/api/messages).¶

<Note>¶
This feature is [Zero Data Retention (ZDR)](/docs/en/build-with-claude/zero-data-retention) eligible. When your organization has a ZDR arrangement, data sent through this feature is not stored after the API response is returned.¶
</Note>¶

## Basic request and response¶

<CodeGroup>¶
```bash Shell¶
#!/bin/sh¶
curl https://api.anthropic.com/v1/messages \¶
--header "x-api-key: $ANTHROPIC_API_KEY" \¶
--header "anthropic-version: 2023-06-01" \¶
--header "content-type: application/json" \¶
--data \¶
'{¶
"model": "claude-opus-4-6",¶
"max_tokens": 1024,¶
"messages": [¶
{"role": "user", "content": "Hello, Claude"}¶
]¶
}'¶
```¶

```python Python¶
import anthropic¶

message = anthropic.Anthropic().messages.create(¶
model="claude-opus-4-6",¶
max_tokens=1024,¶
messages=[{"role": "user", "content": "Hello, Claude"}],¶
)¶
print(message)¶
```¶

```typescript TypeScript¶
import Anthropic from "@anthropic-ai/sdk";¶

const anthropic = new Anthropic();¶

const message = await anthropic.messages.create({¶
model: "claude-opus-4-6",¶
max_tokens: 1024,¶
messages: [
{ role: "user", content: "Hello, Claude" }
]¶
});¶
console.log(message);¶
```¶
</CodeGroup>¶

```json JSON¶
{¶
"id": "msg_01XFDUDYJgAACzvnptvVoYEL",¶
"type": "message",¶
"role": "assistant",¶
"content": [¶
{¶
"type": "text",¶
"text": "Hello!"¶
}¶
],¶
"model": "claude-opus-4-6",¶
"stop_reason": "end_turn",¶
"stop_sequence": null,¶
"usage": {¶
"input_tokens": 12,¶
"output_tokens": 6¶
}¶
}¶
```¶

## Multiple conversational turns¶

The Messages API is stateless, which means that you always send the full conversational history to the API. You can use this pattern to build up a conversation over time. Earlier conversational turns don't necessarily need to actually originate from Claude. You can use synthetic `assistant` messages.¶

<CodeGroup>¶
```bash Shell¶
#!/bin/sh¶
curl https://api.anthropic.com/v1/messages \¶
--header "x-api-key: $ANTHROPIC_API_KEY" \¶
--header "anthropic-version: 2023-06-01" \¶
--header "content-type: application/json" \¶
--data \¶
'{¶
"model": "claude-opus-4-6",¶
"max_tokens": 1024,¶
"messages": [¶
{"role": "user", "content": "Hello, Claude"},¶
{"role": "assistant", "content": "Hello!"},¶
{"role": "user", "content": "Can you describe LLMs to me?"}¶

]¶
}'¶
```¶

```python Python¶
import anthropic¶

message = anthropic.Anthropic().messages.create(¶
model="claude-opus-4-6",¶
max_tokens=1024,¶
messages=[¶
{"role": "user", "content": "Hello, Claude"},¶
{"role": "assistant", "content": "Hello!"},¶
{"role": "user", "content": "Can you describe LLMs to me?"},¶
],¶
)¶
print(message)¶
```¶

```typescript TypeScript¶
import Anthropic from "@anthropic-ai/sdk";¶

const anthropic = new Anthropic();¶

await anthropic.messages.create({¶
model: "claude-opus-4-6",¶
max_tokens: 1024,¶
messages: [¶
{ role: "user", content: "Hello, Claude" },¶
{ role: "assistant", content: "Hello!" },¶
{ role: "user", content: "Can you describe LLMs to me?" }¶
]¶
});¶
```¶
</CodeGroup>¶

```json JSON¶
{¶
 "id": "msg_018gCsTGsXkYJVqYPxTgDHBU",¶
 "type": "message",¶
 "role": "assistant",¶
 "content": [¶
 {¶
 "type": "text",¶
 "text": "Sure, I'd be happy to provide..."¶
 }¶
 ],¶
 "stop_reason": "end_turn",¶
 "stop_sequence": null,¶
 "usage": {¶
 "input_tokens": 30,¶
 "output_tokens": 309¶
 }¶
}¶
```¶

## Putting words in Claude's mouth¶

You can pre-fill part of Claude's response in the last position of the input messages list. This can be used to shape Claude's response. The example below uses `"max_tokens": 1` to get a single multiple choice answer from Claude.¶

<CodeGroup>¶
```bash Shell¶
#!/bin/sh¶
curl https://api.anthropic.com/v1/messages \¶
--header "x-api-key: $ANTHROPIC_API_KEY" \¶
--header "anthropic-version: 2023-06-01" \¶
--header "content-type: application/json" \¶
--data \¶
'{¶
"model": "claude-opus-4-6",¶
"max_tokens": 1,¶
"messages": [¶
{"role": "user", "content": "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae"},¶
{"role": "assistant", "content": "The answer is ("}¶
]¶
}'¶
```¶

```python Python¶
import anthropic¶

message = anthropic.Anthropic().messages.create(¶
model="claude-opus-4-6",¶
max_tokens=1,¶
messages=[¶
{¶
"role": "user",¶
"content": "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae",¶
},¶
{"role": "assistant", "content": "The answer is ("},¶
],¶
)¶
print(message)¶
```¶

```typescript TypeScript¶
import Anthropic from "@anthropic-ai/sdk";¶

const anthropic = new Anthropic();¶

const message = await anthropic.messages.create({¶
model: "claude-opus-4-6",¶
max_tokens: 1,¶
messages: [¶
{
 role: "user",
 content: "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae"
 },¶
{ role: "assistant", content: "The answer is (" }¶
]¶
});¶
console.log(message);¶
```¶
</CodeGroup>¶

```json JSON¶
{¶
"id": "msg_01Q8Faay6S7QPTvEUUQARt7h",¶
"type": "message",¶
"role": "assistant",¶
"content": [¶
{¶
"type": "text",¶
"text": "C"¶
}¶
],¶
"model": "claude-opus-4-6",¶
"stop_reason": "max_tokens",¶
"stop_sequence": null,¶
"usage": {¶
"input_tokens": 42,¶
"output_tokens": 1¶
}¶
}¶
```¶

<Warning>¶
Prefilling is deprecated and not supported on Claude Opus 4.6, Claude Sonnet 4.6, and Claude Sonnet 4.5. Use [structured outputs](/docs/en/build-with-claude/structured-outputs) or system prompt instructions instead.¶
</Warning>¶

## Vision¶

Claude can read both text and images in requests. Both `base64` and `url` source types are supported for images, along with the `image/jpeg`, `image/png`, `image/gif`, and `image/webp` media types. See the [vision guide](/docs/en/build-with-claude/vision) for more details.¶

<CodeGroup>¶
```bash Shell¶
#!/bin/sh¶

# Option 1: Base64-encoded image¶
IMAGE_URL="https://upload.wikimedia.org/wikipedia/commons/a/a7/Camponotus_flavomarginatus_ant.jpg"¶
IMAGE_MEDIA_TYPE="image/jpeg"¶
IMAGE_BASE64=$(curl "$IMAGE_URL" | base64)¶

curl https://api.anthropic.com/v1/messages \¶
--header "x-api-key: $ANTHROPIC_API_KEY" \¶
--header "anthropic-version: 2023-06-01" \¶
--header "content-type: application/json" \¶
--data \¶
'{¶
"model": "claude-opus-4-6",¶
"max_tokens": 1024,¶
"messages": [¶
{"role": "user", "content": [¶
{"type": "image", "source": {¶
"type": "base64",¶
"media_type": "'$IMAGE_MEDIA_TYPE'",¶
"data": "'$IMAGE_BASE64'"¶
}},¶
{"type": "text", "text": "What is in the above image?"}¶
]}¶
]¶
}'¶

# Option 2: URL-referenced image¶
curl https://api.anthropic.com/v1/messages \¶
--header "x-api-key: $ANTHROPIC_API_KEY" \¶
--header "anthropic-version: 2023-06-01" \¶
--header "content-type: application/json" \¶
--data \¶
'{¶
"model": "claude-opus-4-6",¶
"max_tokens": 1024,¶
"messages": [¶
{"role": "user", "content": [¶
{"type": "image", "source": {¶
"type": "url",¶
"url": "https://upload.wikimedia.org/wikipedia/commons/a/a7/Camponotus_flavomarginatus_ant.jpg"¶
}},¶
{"type": "text", "text": "What is in the above image?"}¶
]}¶
]¶
}'¶
```¶

```python Python¶
import anthropic¶
import base64¶
import httpx¶

# Option 1: Base64-encoded image¶
image_url = "https://upload.wikimedia.org/wikipedia/commons/a/a7/Camponotus_flavomarginatus_ant.jpg"¶
image_media_type = "image/jpeg"¶
image_data = base64.standard_b64encode(httpx.get(image_url).content).decode("utf-8")¶

message = anthropic.Anthropic().messages.create(¶
model="claude-opus-4-6",¶
max_tokens=1024,¶
messages=[¶
{¶
"role": "user",¶
"content": [¶
{¶
"type": "image",¶
"source": {¶
"type": "base64",¶
"media_type": image_media_type,¶
"data": image_data,¶
},¶
},¶
{"type": "text", "text": "What is in the above image?"},¶
],¶
}¶
],¶
)¶
print(message)¶

# Option 2: URL-referenced image¶
message_from_url = anthropic.Anthropic().messages.create(¶
model="claude-opus-4-6",¶
max_tokens=1024,¶
messages=[¶
{¶
"role": "user",¶
"content": [¶
{¶
"type": "image",¶
"source": {¶
"type": "url",¶
"url": "https://upload.wikimedia.org/wikipedia/commons/a/a7/Camponotus_flavomarginatus_ant.jpg",¶
},¶
},¶
{"type": "text", "text": "What is in the above image?"},¶
],¶
}¶
],¶
)¶
print(message_from_url)¶
```¶

```typescript TypeScript¶
import Anthropic from "@anthropic-ai/sdk";¶

const anthropic = new Anthropic();¶

// Option 1: Base64-encoded image¶
const image_url =
 "https://upload.wikimedia.org/wikipedia/commons/a/a7/Camponotus_flavomarginatus_ant.jpg";¶
const image_media_type = "image/jpeg";¶
const image_array_buffer = await ((await fetch(image_url)).arrayBuffer());¶
const image_data = Buffer.from(image_array_buffer).toString("base64");¶

const message = await anthropic.messages.create({¶
model: "claude-opus-4-6",¶
max_tokens: 1024,¶
messages: [¶
{¶
role: "user",¶
content: [¶
{¶
type: "image",¶
source: {¶
type: "base64",¶
media_type: image_media_type,¶
data: image_data¶
}¶
},¶
{¶
type: "text",¶
text: "What is in the above image?"¶
}¶
]¶
}¶
]¶
});¶
console.log(message);¶

// Option 2: URL-referenced image¶
const messageFromUrl = await anthropic.messages.create({¶
model: "claude-opus-4-6",¶
max_tokens: 1024,¶
messages: [¶
{¶
role: "user",¶
content: [¶
{¶
type: "image",¶
source: {¶
type: "url",¶
url: "https://upload.wikimedia.org/wikipedia/commons/a/a7/Camponotus_flavomarginatus_ant.jpg"¶
}¶
},¶
{¶
type: "text",¶
text: "What is in the above image?"¶
}¶
]¶
}¶
]¶
});¶
console.log(messageFromUrl);¶
```¶
</CodeGroup>¶

```json JSON¶
{¶
"id": "msg_01EcyWo6m4hyW8KHs2y2pei5",¶
"type": "message",¶
"role": "assistant",¶
"content": [¶
{¶
"type": "text",¶
"text": "This image shows an ant, specifically a close-up view of an ant. The ant is shown in detail, with its distinct head, antennae, and legs clearly visible. The image is focused on capturing the intricate details and features of the ant, likely taken with a macro lens to get an extreme close-up perspective."¶
}¶
],¶
"model": "claude-opus-4-6",¶
"stop_reason": "end_turn",¶
"stop_sequence": null,¶
"usage": {¶
"input_tokens": 1551,¶
"output_tokens": 71¶
}¶
}¶
```¶

## Tool use and computer use¶

See the [tool use guide](/docs/en/agents-and-tools/tool-use/overview) for examples of how to use tools with the Messages API.¶
See the [computer use guide](/docs/en/agents-and-tools/tool-use/computer-use-tool) for examples of how to control desktop computer environments with the Messages API.¶
For guaranteed JSON output, see [Structured Outputs](/docs/en/build-with-claude/structured-outputs).

Unified Diff

--- a/build-with-claude/working-with-messages.md
+++ b/build-with-claude/working-with-messages.md
@@ -48,9 +48,7 @@
   const message = await anthropic.messages.create({
     model: "claude-opus-4-6",
     max_tokens: 1024,
-    messages: [
-      { role: "user", content: "Hello, Claude" }
-    ]
+    messages: [{ role: "user", content: "Hello, Claude" }]
   });
   console.log(message);
   ```
@@ -135,21 +133,21 @@
 
 ```json JSON
 {
-    "id": "msg_018gCsTGsXkYJVqYPxTgDHBU",
-    "type": "message",
-    "role": "assistant",
-    "content": [
-        {
-            "type": "text",
-            "text": "Sure, I'd be happy to provide..."
-        }
-    ],
-    "stop_reason": "end_turn",
-    "stop_sequence": null,
-    "usage": {
-      "input_tokens": 30,
-      "output_tokens": 309
+  "id": "msg_018gCsTGsXkYJVqYPxTgDHBU",
+  "type": "message",
+  "role": "assistant",
+  "content": [
+    {
+      "type": "text",
+      "text": "Sure, I'd be happy to provide..."
     }
+  ],
+  "stop_reason": "end_turn",
+  "stop_sequence": null,
+  "usage": {
+    "input_tokens": 30,
+    "output_tokens": 309
+  }
 }
 ```
 
@@ -201,7 +199,10 @@
     model: "claude-opus-4-6",
     max_tokens: 1,
     messages: [
-      { role: "user", content: "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae" },
+      {
+        role: "user",
+        content: "What is latin for Ant? (A) Apoidea, (B) Rhopalocera, (C) Formicidae"
+      },
       { role: "assistant", content: "The answer is (" }
     ]
   });
@@ -349,9 +350,10 @@
   const anthropic = new Anthropic();
 
   // Option 1: Base64-encoded image
-  const image_url = "https://upload.wikimedia.org/wikipedia/commons/a/a7/Camponotus_flavomarginatus_ant.jpg";
+  const image_url =
+    "https://upload.wikimedia.org/wikipedia/commons/a/a7/Camponotus_flavomarginatus_ant.jpg";
   const image_media_type = "image/jpeg";
-  const image_array_buffer = await ((await fetch(image_url)).arrayBuffer());
+  const image_array_buffer = await (await fetch(image_url)).arrayBuffer();
   const image_data = Buffer.from(image_array_buffer).toString("base64");
 
   const message = await anthropic.messages.create({