← Back to daily report

build-with-claude/claude-on-vertex-ai.md

Changed on 2026-02-06 18:20:36 EST

+95 lines added
-2 lines removed
Visual Diff
# Claude on Vertex AI¶

Anthropic's Claude models are now generally available through [Vertex AI](https://cloud.google.com/vertex-ai).¶

---¶

The Vertex API for accessing Claude is nearly-identical to the [Messages API](/docs/en/api/messages) and supports all of the same options, with two key differences:¶

* In Vertex, `model` is not passed in the request body. Instead, it is specified in the Google Cloud endpoint URL.¶
* In Vertex, `anthropic_version` is passed in the request body (rather than as a header), and must be set to the value `vertex-2023-10-16`.¶

Vertex is also supported by Anthropic's official [client SDKs](/docs/en/api/client-sdks). This guide will walk you through the process of making a request to Claude on Vertex AI in either Python or TypeScript.¶

Note that this guide assumes you have already have a GCP project that is able to use Vertex AI. See [using the Claude 3 models from Anthropic](https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/use-claude) for more information on the setup required, as well as a full walkthrough.¶

## Install an SDK for accessing Vertex AI¶

First, install Anthropic's [client SDK](/docs/en/api/client-sdks) for your language of choice.¶

<
CodeGroup>¶
```python Python¶
Tabs>¶
<Tab title="Python">¶
```bash¶
pip install -U google-cloud-aiplatform "anthropic[vertex]"¶
```¶

```typescript TypeScript¶
npm install @anthropic-ai/vertex-sdk¶
```¶
</CodeGroup
</Tab>¶

<Tab title="TypeScript">¶
```bash¶
npm install @anthropic-ai/vertex-sdk¶
```¶
</Tab>¶

<Tab title="Java">¶
<CodeGroup>¶
```groovy Gradle¶
implementation("com.anthropic:anthropic-java-vertex:2.+")¶
```¶

```xml Maven¶
<dependency>¶
<groupId>com.anthropic</groupId>¶
<artifactId>anthropic-java-vertex</artifactId>¶
<version>2.13.0</version>¶
</dependency>¶
```¶
</CodeGroup>¶
</Tab>¶

<Tab title="Go">¶
```bash¶
go get github.com/anthropics/anthropic-sdk-go¶
```¶
</Tab>¶
</Tabs
>¶

## Accessing Vertex AI¶

### Model Availability¶

Note that Anthropic model availability varies by region. Search for "Claude" in the [Vertex AI Model Garden](https://cloud.google.com/model-garden) or go to [Use Claude 3](https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/use-claude) for the latest information.¶

#### API model IDs¶

| Model | Vertex AI API model ID |¶
| ------------------------------ | ------------------------ |¶
| Claude Opus 4.6 | claude-opus-4-6 |¶
| Claude Sonnet 4.5 | claude-sonnet-4-5@20250929 |¶
| Claude Sonnet 4 | claude-sonnet-4@20250514 |¶
| Claude Sonnet 3.7 <Tooltip tooltipContent="Deprecated as of October 28, 2025.">⚠️</Tooltip> | claude-3-7-sonnet@20250219 |¶
| Claude Opus 4.5 | claude-opus-4-5@20251101 |¶
| Claude Opus 4.1 | claude-opus-4-1@20250805 |¶
| Claude Opus 4 | claude-opus-4@20250514 |¶
| Claude Haiku 4.5 | claude-haiku-4-5@20251001 |¶
| Claude Haiku 3.5 <Tooltip tooltipContent="Deprecated as of December 19, 2025.">⚠️</Tooltip> | claude-3-5-haiku@20241022 |¶
| Claude Haiku 3 | claude-3-haiku@20240307 |¶

### Making requests¶

Before running requests you may need to run `gcloud auth application-default login` to authenticate with GCP.¶

The following examples shows how to generate text from Claude on Vertex AI:¶
<CodeGroup>¶

```python Python¶
from anthropic import AnthropicVertex¶

project_id = "MY_PROJECT_ID"¶
region = "global"¶

client = AnthropicVertex(project_id=project_id, region=region)¶

message = client.messages.create(¶
model="claude-opus-4-6",¶
max_tokens=100,¶
messages=[¶
{¶
"role": "user",¶
"content": "Hey Claude!",¶
}¶
],¶
)¶
print(message)¶
```¶

```typescript TypeScript¶
import { AnthropicVertex } from '@anthropic-ai/vertex-sdk';¶

const projectId = 'MY_PROJECT_ID';¶
const region = 'global';¶

// Goes through the standard `google-auth-library` flow.¶
const client = new AnthropicVertex({¶
projectId,¶
region,¶
});¶

async function main() {¶
const result = await client.messages.create({¶
model: 'claude-opus-4-6',¶
max_tokens: 100,¶
messages: [¶
{¶
role: 'user',¶
content: 'Hey Claude!',¶
},¶
],¶
});¶
console.log(JSON.stringify(result, null, 2));¶
}¶

main();¶
```¶

```java Java¶
import com.anthropic.client.AnthropicClient;¶
import com.anthropic.client.okhttp.AnthropicOkHttpClient;¶
import com.anthropic.vertex.backends.VertexBackend;¶
import com.anthropic.models.messages.Message;¶
import com.anthropic.models.messages.MessageCreateParams;¶

public class VertexExample {¶
public static void main(String[] args) {¶
// Uses default Google Cloud credentials¶
AnthropicClient client = AnthropicOkHttpClient.builder()¶
.backend(VertexBackend.fromEnv())¶
.build();¶

Message message = client.messages().create(MessageCreateParams.builder()¶
.model("claude-opus-4-6")¶
.maxTokens(100)¶
.addUserMessage("Hey Claude!")¶
.build());¶

System.out.println(message);¶
}¶
}¶
```¶

```go Go¶
package main¶

import (¶
"context"¶
"fmt"¶

"github.com/anthropics/anthropic-sdk-go"¶
"github.com/anthropics/anthropic-sdk-go/vertex"¶
)¶

func main() {¶
// Uses default Google Cloud credentials¶
client := anthropic.NewClient(¶
vertex.WithGoogleAuth(context.Background(), "global", "MY_PROJECT_ID"),¶
)¶

message, err := client.Messages.New(context.Background(), anthropic.MessageNewParams{¶
Model: "claude-opus-4-6",¶
MaxTokens: 100,¶
Messages: []anthropic.MessageParam{¶
anthropic.NewUserMessage(anthropic.NewTextBlock("Hey Claude!")),¶
},¶
})¶
if err != nil {¶
panic(err)¶
}¶
fmt.Printf("%+v\n", message)¶
}¶
```¶

```bash Shell¶
MODEL_ID=claude-opus-4-6¶
LOCATION=global¶
PROJECT_ID=MY_PROJECT_ID¶

curl \¶
-X POST \¶
-H "Authorization: Bearer $(gcloud auth print-access-token)" \¶
-H "Content-Type: application/json" \¶
https://$LOCATION-aiplatform.googleapis.com/v1/projects/${PROJECT_ID}/locations/${LOCATION}/publishers/anthropic/models/${MODEL_ID}:streamRawPredict -d \¶
'{¶
"anthropic_version": "vertex-2023-10-16",¶
"messages": [{¶
"role": "user",¶
"content": "Hey Claude!"¶
}],¶
"max_tokens": 100,¶
}'¶
```¶
</CodeGroup>¶

See our [client SDKs](/docs/en/api/client-sdks) and the official [Vertex AI docs](https://cloud.google.com/vertex-ai/docs) for more details.¶

Claude is also available through [Amazon Bedrock](/docs/en/build-with-claude/claude-on-amazon-bedrock) and [Microsoft Foundry](/docs/en/build-with-claude/claude-in-microsoft-foundry).¶

## Activity logging¶

Vertex provides a [request-response logging service](https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/request-response-logging) that allows customers to log the prompts and completions associated with your usage. ¶

Anthropic recommends that you log your activity on at least a 30-day rolling basis in order to understand your activity and investigate any potential misuse. ¶

<Note>¶
Turning on this service does not give Google or Anthropic any access to your content.¶
</Note>¶

## Feature support¶
You can find all the features currently supported on Vertex [here](/docs/en/api/overview).¶

## Global vs regional endpoints¶

Starting with **Claude Sonnet 4.5 and all future models**, Google Vertex AI offers two endpoint types:¶

- **Global endpoints**: Dynamic routing for maximum availability¶
- **Regional endpoints**: Guaranteed data routing through specific geographic regions¶

Regional endpoints include a 10% pricing premium over global endpoints.¶

<Note>¶
This applies to Claude Sonnet 4.5 and future models only. Older models (Claude Sonnet 4, Opus 4, and earlier) maintain their existing pricing structures.¶
</Note>¶

### When to use each option¶

**Global endpoints (recommended):**¶
- Provide maximum availability and uptime¶
- Dynamically route requests to regions with available capacity¶
- No pricing premium¶
- Best for applications where data residency is flexible¶
- Only supports pay-as-you-go traffic (provisioned throughput requires regional endpoints)¶

**Regional endpoints:**¶
- Route traffic through specific geographic regions¶
- Required for data residency and compliance requirements¶
- Support both pay-as-you-go and provisioned throughput¶
- 10% pricing premium reflects infrastructure costs for dedicated regional capacity¶

### Implementation¶

**Using global endpoints (recommended):**¶

Set the `region` parameter to `"global"` when initializing the client:¶

<CodeGroup>¶
```python Python¶
from anthropic import AnthropicVertex¶

project_id = "MY_PROJECT_ID"¶
region = "global"¶

client = AnthropicVertex(project_id=project_id, region=region)¶

message = client.messages.create(¶
model="claude-opus-4-6",¶
max_tokens=100,¶
messages=[¶
{¶
"role": "user",¶
"content": "Hey Claude!",¶
}¶
],¶
)¶
print(message)¶
```¶

```typescript TypeScript¶
import { AnthropicVertex } from '@anthropic-ai/vertex-sdk';¶

const projectId = 'MY_PROJECT_ID';¶
const region = 'global';¶

const client = new AnthropicVertex({¶
projectId,¶
region,¶
});¶

const result = await client.messages.create({¶
model: 'claude-opus-4-6',¶
max_tokens: 100,¶
messages: [¶
{¶
role: 'user',¶
content: 'Hey Claude!',¶
},¶
],¶
});¶
```¶
</CodeGroup>¶

**Using regional endpoints:**¶

Specify a specific region like `"us-east1"` or `"europe-west1"`:¶

<CodeGroup>¶
```python Python¶
from anthropic import AnthropicVertex¶

project_id = "MY_PROJECT_ID"¶
region = "us-east1" # Specify a specific region¶

client = AnthropicVertex(project_id=project_id, region=region)¶

message = client.messages.create(¶
model="claude-opus-4-6",¶
max_tokens=100,¶
messages=[¶
{¶
"role": "user",¶
"content": "Hey Claude!",¶
}¶
],¶
)¶
print(message)¶
```¶

```typescript TypeScript¶
import { AnthropicVertex } from '@anthropic-ai/vertex-sdk';¶

const projectId = 'MY_PROJECT_ID';¶
const region = 'us-east1'; // Specify a specific region¶

const client = new AnthropicVertex({¶
projectId,¶
region,¶
});¶

const result = await client.messages.create({¶
model: 'claude-opus-4-6',¶
max_tokens: 100,¶
messages: [¶
{¶
role: 'user',¶
content: 'Hey Claude!',¶
},¶
],¶
});

```java Java¶
import com.anthropic.client.AnthropicClient;¶
import com.anthropic.client.okhttp.AnthropicOkHttpClient;¶
import com.anthropic.vertex.backends.VertexBackend;¶
import com.anthropic.models.messages.MessageCreateParams;¶

// Uses default Google Cloud credentials¶
AnthropicClient client = AnthropicOkHttpClient.builder()¶
.backend(VertexBackend.fromEnv())¶
.build();¶

var message = client.messages().create(MessageCreateParams.builder()¶
.model("claude-opus-4-6")¶
.maxTokens(100)¶
.addUserMessage("Hey Claude!")¶
.build());¶
```¶

```go Go¶
import (¶
"context"¶

"github.com/anthropics/anthropic-sdk-go"¶
"github.com/anthropics/anthropic-sdk-go/vertex"¶
)¶

// Uses default Google Cloud credentials¶
client := anthropic.NewClient(¶
vertex.WithGoogleAuth(context.Background(), "global", "MY_PROJECT_ID"),¶
)¶

message, _ := client.Messages.New(context.Background(), anthropic.MessageNewParams{¶
Model: "claude-opus-4-6",¶
MaxTokens: 100,¶
Messages: []anthropic.MessageParam{¶
anthropic.NewUserMessage(anthropic.NewTextBlock("Hey Claude!")),¶
},¶
})¶
```¶
</CodeGroup>¶

**Using regional endpoints:**¶

Specify a specific region like `"us-east1"` or `"europe-west1"`:¶

<CodeGroup>¶
```python Python¶
from anthropic import AnthropicVertex¶

project_id = "MY_PROJECT_ID"¶
region = "us-east1" # Specify a specific region¶

client = AnthropicVertex(project_id=project_id, region=region)¶

message = client.messages.create(¶
model="claude-opus-4-6",¶
max_tokens=100,¶
messages=[¶
{¶
"role": "user",¶
"content": "Hey Claude!",¶
}¶
],¶
)¶
print(message)¶
```¶

```typescript TypeScript¶
import { AnthropicVertex } from '@anthropic-ai/vertex-sdk';¶

const projectId = 'MY_PROJECT_ID';¶
const region = 'us-east1'; // Specify a specific region¶

const client = new AnthropicVertex({¶
projectId,¶
region,¶
});¶

const result = await client.messages.create({¶
model: 'claude-opus-4-6',¶
max_tokens: 100,¶
messages: [¶
{¶
role: 'user',¶
content: 'Hey Claude!',¶
},¶
],¶
});¶
```¶

```java Java¶
import com.anthropic.client.AnthropicClient;¶
import com.anthropic.client.okhttp.AnthropicOkHttpClient;¶
import com.anthropic.vertex.backends.VertexBackend;¶
import com.anthropic.models.messages.MessageCreateParams;¶

// Uses default Google Cloud credentials with specific region¶
AnthropicClient client = AnthropicOkHttpClient.builder()¶
.backend(VertexBackend.builder()¶
.region("us-east1") // Specify a specific region¶
.project("MY_PROJECT_ID")¶
.build())¶
.build();¶

var message = client.messages().create(MessageCreateParams.builder()¶
.model("claude-opus-4-6")¶
.maxTokens(100)¶
.addUserMessage("Hey Claude!")¶
.build());¶
```¶

```go Go¶
import (¶
"context"¶

"github.com/anthropics/anthropic-sdk-go"¶
"github.com/anthropics/anthropic-sdk-go/vertex"¶
)¶

// Specify a specific region¶
client := anthropic.NewClient(¶
vertex.WithGoogleAuth(context.Background(), "us-east1", "MY_PROJECT_ID"),¶
)¶

message, _ := client.Messages.New(context.Background(), anthropic.MessageNewParams{¶
Model: "claude-opus-4-6",¶
MaxTokens: 100,¶
Messages: []anthropic.MessageParam{¶
anthropic.NewUserMessage(anthropic.NewTextBlock("Hey Claude!")),¶
},¶
})

```¶
</CodeGroup>¶

### Additional resources¶

- **Google Vertex AI pricing:** [cloud.google.com/vertex-ai/generative-ai/pricing](https://cloud.google.com/vertex-ai/generative-ai/pricing)¶
- **Claude models documentation:** [Claude on Vertex AI](https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude)¶
- **Google blog post:** [Global endpoint for Claude models](https://cloud.google.com/blog/products/ai-machine-learning/global-endpoint-for-claude-models-generally-available-on-vertex-ai)¶
- **Anthropic pricing details:** [Pricing documentation](/docs/en/about-claude/pricing#third-party-platform-pricing)

Unified Diff

--- a/build-with-claude/claude-on-vertex-ai.md
+++ b/build-with-claude/claude-on-vertex-ai.md
@@ -17,15 +17,41 @@
 
 First, install Anthropic's [client SDK](/docs/en/api/client-sdks) for your language of choice.
 
+<Tabs>
+<Tab title="Python">
+```bash
+pip install -U google-cloud-aiplatform "anthropic[vertex]"
+```
+</Tab>
+
+<Tab title="TypeScript">
+```bash
+npm install @anthropic-ai/vertex-sdk
+```
+</Tab>
+
+<Tab title="Java">
 <CodeGroup>
-  ```python Python
-  pip install -U google-cloud-aiplatform "anthropic[vertex]"
-  ```
-
-  ```typescript TypeScript
-  npm install @anthropic-ai/vertex-sdk
-  ```
+```groovy Gradle
+implementation("com.anthropic:anthropic-java-vertex:2.+")
+```
+
+```xml Maven
+<dependency>
+    <groupId>com.anthropic</groupId>
+    <artifactId>anthropic-java-vertex</artifactId>
+    <version>2.13.0</version>
+</dependency>
+```
 </CodeGroup>
+</Tab>
+
+<Tab title="Go">
+```bash
+go get github.com/anthropics/anthropic-sdk-go
+```
+</Tab>
+</Tabs>
 
 ## Accessing Vertex AI
 
@@ -103,6 +129,62 @@
   }
 
   main();
+  ```
+
+  ```java Java
+  import com.anthropic.client.AnthropicClient;
+  import com.anthropic.client.okhttp.AnthropicOkHttpClient;
+  import com.anthropic.vertex.backends.VertexBackend;
+  import com.anthropic.models.messages.Message;
+  import com.anthropic.models.messages.MessageCreateParams;
+
+  public class VertexExample {
+      public static void main(String[] args) {
+          // Uses default Google Cloud credentials
+          AnthropicClient client = AnthropicOkHttpClient.builder()
+              .backend(VertexBackend.fromEnv())
+              .build();
+
+          Message message = client.messages().create(MessageCreateParams.builder()
+              .model("claude-opus-4-6")
+              .maxTokens(100)
+              .addUserMessage("Hey Claude!")
+              .build());
+
+          System.out.println(message);
+      }
+  }
+  ```
+
+  ```go Go
+  package main
+
+  import (
+      "context"
+      "fmt"
+
+      "github.com/anthropics/anthropic-sdk-go"
+      "github.com/anthropics/anthropic-sdk-go/vertex"
+  )
+
+  func main() {
+      // Uses default Google Cloud credentials
+      client := anthropic.NewClient(
+          vertex.WithGoogleAuth(context.Background(), "global", "MY_PROJECT_ID"),
+      )
+
+      message, err := client.Messages.New(context.Background(), anthropic.MessageNewParams{
+          Model:     "claude-opus-4-6",
+          MaxTokens: 100,
+          Messages: []anthropic.MessageParam{
+              anthropic.NewUserMessage(anthropic.NewTextBlock("Hey Claude!")),
+          },
+      })
+      if err != nil {
+          panic(err)
+      }
+      fmt.Printf("%+v\n", message)
+  }
   ```
 
   ```bash Shell
@@ -128,6 +210,8 @@
 
 See our [client SDKs](/docs/en/api/client-sdks) and the official [Vertex AI docs](https://cloud.google.com/vertex-ai/docs) for more details.
 
+Claude is also available through [Amazon Bedrock](/docs/en/build-with-claude/claude-on-amazon-bedrock) and [Microsoft Foundry](/docs/en/build-with-claude/claude-in-microsoft-foundry).
+
 ## Activity logging
 
 Vertex provides a [request-response logging service](https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/request-response-logging) that allows customers to log the prompts and completions associated with your usage. 
@@ -219,6 +303,46 @@
   ],
 });
 ```
+
+```java Java
+import com.anthropic.client.AnthropicClient;
+import com.anthropic.client.okhttp.AnthropicOkHttpClient;
+import com.anthropic.vertex.backends.VertexBackend;
+import com.anthropic.models.messages.MessageCreateParams;
+
+// Uses default Google Cloud credentials
+AnthropicClient client = AnthropicOkHttpClient.builder()
+    .backend(VertexBackend.fromEnv())
+    .build();
+
+var message = client.messages().create(MessageCreateParams.builder()
+    .model("claude-opus-4-6")
+    .maxTokens(100)
+    .addUserMessage("Hey Claude!")
+    .build());
+```
+
+```go Go
+import (
+    "context"
+
+    "github.com/anthropics/anthropic-sdk-go"
+    "github.com/anthropics/anthropic-sdk-go/vertex"
+)
+
+// Uses default Google Cloud credentials
+client := anthropic.NewClient(
+    vertex.WithGoogleAuth(context.Background(), "global", "MY_PROJECT_ID"),
+)
+
+message, _ := client.Messages.New(context.Background(), anthropic.MessageNewParams{
+    Model:     "claude-opus-4-6",
+    MaxTokens: 100,
+    Messages: []anthropic.MessageParam{
+        anthropic.NewUserMessage(anthropic.NewTextBlock("Hey Claude!")),
+    },
+})
+```
 </CodeGroup>
 
 **Using regional endpoints:**
@@ -269,6 +393,49 @@
   ],
 });
 ```
+
+```java Java
+import com.anthropic.client.AnthropicClient;
+import com.anthropic.client.okhttp.AnthropicOkHttpClient;
+import com.anthropic.vertex.backends.VertexBackend;
+import com.anthropic.models.messages.MessageCreateParams;
+
+// Uses default Google Cloud credentials with specific region
+AnthropicClient client = AnthropicOkHttpClient.builder()
+    .backend(VertexBackend.builder()
+        .region("us-east1")  // Specify a specific region
+        .project("MY_PROJECT_ID")
+        .build())
+    .build();
+
+var message = client.messages().create(MessageCreateParams.builder()
+    .model("claude-opus-4-6")
+    .maxTokens(100)
+    .addUserMessage("Hey Claude!")
+    .build());
+```
+
+```go Go
+import (
+    "context"
+
+    "github.com/anthropics/anthropic-sdk-go"
+    "github.com/anthropics/anthropic-sdk-go/vertex"
+)
+
+// Specify a specific region
+client := anthropic.NewClient(
+    vertex.WithGoogleAuth(context.Background(), "us-east1", "MY_PROJECT_ID"),
+)
+
+message, _ := client.Messages.New(context.Background(), anthropic.MessageNewParams{
+    Model:     "claude-opus-4-6",
+    MaxTokens: 100,
+    Messages: []anthropic.MessageParam{
+        anthropic.NewUserMessage(anthropic.NewTextBlock("Hey Claude!")),
+    },
+})
+```
 </CodeGroup>
 
 ### Additional resources