Anthropic Provider

The Anthropic provider gives you access to Claude models within the AI SDK. It supports Claude 3 models with capabilities for chat, image understanding, and tool use.

Basic Usage

The simplest way to use Anthropic models is with the anthropic function:


from ai_sdk import generate_text
from ai_sdk.anthropic import anthropic
 
response = generate_text(
    model=anthropic("claude-3-opus-20240229"),
    prompt="Hello, world!"
)

Supported Models

The following table shows all supported Claude models and their capabilities:

Model	Chat	Images	Tool Calls
claude-3-7-sonnet-20250219	✅	✅	✅
claude-3-5-sonnet-20241022	✅	✅	✅
claude-3-5-sonnet-20240620	✅	✅	✅
claude-3-5-haiku-20241022	✅	❌	✅
claude-3-opus-20240229	✅	✅	✅
claude-3-sonnet-20240229	✅	✅	❌
claude-3-haiku-20240307	✅	✅	✅

All models support basic chat functionality. Image and tool call support varies by model.

Configuration

You can configure the Anthropic provider with various settings:


from ai_sdk.anthropic import create_anthropic_provider, AnthropicProviderSettings
 
provider = create_anthropic_provider(
    settings=AnthropicProviderSettings(
        api_key="your-api-key",  # Optional: can also use ANTHROPIC_API_KEY env var
        base_url="https://api.anthropic.com",  # Optional: defaults to Anthropic's API
        headers={"Custom-Header": "value"}  # Optional: additional headers
    )
)
 
model = provider("claude-3-opus-20240229")

Authentication

The provider will look for your API key in the following order:

The api_key parameter in settings
The ANTHROPIC_API_KEY environment variable

⚠️

Never hardcode your API key in your code. Use environment variables instead.

Model Settings

When creating a model, you can customize its behavior:


from ai_sdk.anthropic import anthropic, AnthropicChatSettings
 
model = anthropic(
    "claude-3-opus-20240229", 
    settings=AnthropicChatSettings(
        sendReasoning=True  # Enable reasoning in responses
    )
)

Working with Images

For models that support image input, you can pass images via URL or base64:


response = generate_text(
    model=anthropic("claude-3-opus-20240229"),
    messages=[{
        "role": "user",
        "content": [
            {"type": "text", "text": "What's in this image?"},
            {"type": "image", "image": "https://example.com/image.jpg"}
        ]
    }]
)

Image support is available on specific Claude 3 models only. Check SUPPORTED_IMAGE_MODELS for compatibility.

Using Tool Calls

Tools can be defined using Pydantic models and optional execution functions:


from pydantic import BaseModel
from ai_sdk import generate_text
from typing import Dict
 
class WeatherTool(BaseModel):
    location: str
    units: str = "celsius"
 
# Define tools with optional execution functions
tools: Dict[str, dict] = {
    "weather": {
        "description": "Get the weather in a location",
        "parameters": WeatherTool,
        "execute": lambda args: {
            "temperature": 22,
            "condition": "sunny"
        }
    }
}
 
response = generate_text(
    model=anthropic("claude-3-opus-20240229"),
    prompt="What's the weather in Paris?",
    tools=tools,
    max_steps=2  # Allow multiple steps for tool execution
)
 
# Access results
print(response.text)  # Final response
print(response.tool_calls)  # List of tool calls made
print(response.tool_results)  # List of tool results

The response includes:

text: The final generated text
tool_calls: List of ToolCallPart objects containing:
- tool_call_id: Unique identifier for the call
- tool_name: Name of the tool called
- args: Dictionary of arguments passed
tool_results: List of ToolResultPart objects containing:
- tool_call_id: Matching ID from the call
- tool_name: Name of the tool
- result: The result returned
- is_error: Optional error flag

When max_steps is greater than 1, the model can make tool calls and use the results in its final response.

⚠️

Not all Claude models support tool calls. Check the compatibility table above for supported models.