Python SDK

The Anthropic Python SDK provides convenient access to the Claude API from Python applications. It supports both synchronous and asynchronous operations, streaming, and integrations with Amazon Bedrock, Claude Platform on AWS, Google Cloud, and Microsoft Foundry.

Installation

pip install anthropic

For platform-specific integrations or improved async performance, install with extras:

# For Amazon Bedrock support
pip install "anthropic[bedrock]"

# For Google Cloud support
pip install "anthropic[vertex]"

# For Claude Platform on AWS support
pip install "anthropic[aws]"

# Microsoft Foundry support is included in the base package

# For improved async performance with aiohttp
pip install "anthropic[aiohttp]"

Requirements

Python 3.10 or later is required. If you are upgrading from a 0.x release of the SDK, see the v1 migration guide for the list of breaking changes.

Usage

import os
from anthropic import Anthropic

client = Anthropic(
    # This is the default and can be omitted
    api_key=os.environ.get("ANTHROPIC_API_KEY"),
)

message = client.messages.create(
    max_tokens=1024,
    messages=[
        {
            "role": "user",
            "content": "Hello, Claude",
        }
    ],
    model="claude-opus-5",
)

for block in message.content:
    if block.type == "text":
        print(block.text)

For authentication options including Workload Identity Federation, see Authentication. If your API key is a personal or service account key with access to multiple workspaces, set the workspace ID in the anthropic-workspace-id request header; Select a workspace shows the per-request option for this SDK.

Async usage

import os
import asyncio
from anthropic import AsyncAnthropic

client = AsyncAnthropic(
    api_key=os.environ.get("ANTHROPIC_API_KEY"),
)


async def main() -> None:
    message = await client.messages.create(
        max_tokens=1024,
        messages=[
            {
                "role": "user",
                "content": "Hello, Claude",
            }
        ],
        model="claude-opus-5",
    )
    print(message.content)


asyncio.run(main())

Using aiohttp for better concurrency

For improved async performance, you can use the aiohttp HTTP backend instead of the default httpx2:

import os
import asyncio
from anthropic import AsyncAnthropic, DefaultAioHttpClient


async def main() -> None:
    async with AsyncAnthropic(
        api_key=os.environ.get("ANTHROPIC_API_KEY"),
        http_client=DefaultAioHttpClient(),
    ) as client:
        message = await client.messages.create(
            max_tokens=1024,
            messages=[
                {
                    "role": "user",
                    "content": "Hello, Claude",
                }
            ],
            model="claude-opus-5",
        )
        print(message.content)


asyncio.run(main())

Streaming responses

The SDK provides support for streaming responses using Server-Sent Events (SSE).

client = Anthropic()

stream = client.messages.create(
    max_tokens=1024,
    messages=[
        {
            "role": "user",
            "content": "Hello, Claude",
        }
    ],
    model="claude-opus-5",
    stream=True,
)
for event in stream:
    print(event.type)

The async client uses the exact same interface:

client = AsyncAnthropic()

stream = await client.messages.create(
    max_tokens=1024,
    messages=[
        {
            "role": "user",
            "content": "Hello, Claude",
        }
    ],
    model="claude-opus-5",
    stream=True,
)
async for event in stream:
    print(event.type)

Streaming helpers

The SDK also provides streaming helpers that use context managers and provide access to the accumulated text and the final message:

async def main() -> None:
    async with client.messages.stream(
        max_tokens=1024,
        messages=[
            {
                "role": "user",
                "content": "Say hello there!",
            }
        ],
        model="claude-opus-5",
    ) as stream:
        async for text in stream.text_stream:
            print(text, end="", flush=True)
        print()

        message = await stream.get_final_message()
        print(message.to_json())


asyncio.run(main())

Streaming with client.messages.stream(...) exposes various helpers including accumulation and SDK-specific events.

Alternatively, you can use client.messages.create(..., stream=True) which only returns an iterable of the events in the stream and uses less memory (it doesn't build up a final message object for you).

Token counting

You can see the exact usage for a given request through the usage response property:

message = client.messages.create(...)
print(message.usage)
# Usage(input_tokens=25, output_tokens=13)

You can also count tokens before making a request:

count = client.messages.count_tokens(
    model="claude-opus-5", messages=[{"role": "user", "content": "Hello, world"}]
)
print(count.input_tokens)  # 10

Tool use

This SDK provides support for tool use, also known as function calling. For more details, see Tool use with Claude.

Tool helpers

The SDK provides helpers for defining and running tools as pure Python functions. The @beta_tool decorator generates the tool schema from the function signature and docstring:

import json
from anthropic import Anthropic, beta_tool

client = Anthropic()


@beta_tool
def get_weather(location: str) -> str:
    """Get the weather for a given location.

    Args:
        location: The city and state, for example, San Francisco, CA
    Returns:
        A JSON-encoded string with the location, temperature, and weather condition.
    """
    return json.dumps(
        {
            "location": location,
            "temperature": "68°F",
            "condition": "Sunny",
        }
    )


# Use the tool_runner to automatically handle tool calls
runner = client.beta.messages.tool_runner(
    max_tokens=1024,
    model="claude-opus-5",
    tools=[get_weather],
    messages=[
        {"role": "user", "content": "What is the weather in SF?"},
    ],
)
for message in runner:
    print(message)

On every iteration, an API request is made. If the response includes a call to one of the given tools, the tool is automatically called, and the result is returned directly to the model in the next iteration.

Message batches

This SDK provides support for Batch processing under client.messages.batches.

Creating a batch

Message Batches takes an array of requests, where each object has a custom_id identifier and the same request params as the standard Messages API:

client.messages.batches.create(
    requests=[
        {
            "custom_id": "my-first-request",
            "params": {
                "model": "claude-opus-5",
                "max_tokens": 1024,
                "messages": [{"role": "user", "content": "Hello, world"}],
            },
        },
        {
            "custom_id": "my-second-request",
            "params": {
                "model": "claude-opus-5",
                "max_tokens": 1024,
                "messages": [{"role": "user", "content": "Hi again, friend"}],
            },
        },
    ]
)

Getting results from a batch

Once a Message Batch has been processed, indicated by .processing_status == 'ended', you can access the results with .batches.results():

client = anthropic.Anthropic()
batch_id = "batch_abc123"
result_stream = client.messages.batches.results(batch_id)
for entry in result_stream:
    if entry.result.type == "succeeded":
        print(entry.result.message.content)

File uploads

Request parameters that correspond to file uploads can be passed in many different forms:

• A PathLike object (for example, pathlib.Path)
• A tuple of (filename, content, content_type)
• A BinaryIO file-like object

from pathlib import Path
from anthropic import Anthropic

client = Anthropic()

# Upload using a file path
client.files.upload(
    file=Path("/path/to/file"),
)

# Upload using bytes
client.files.upload(
    file=("file.txt", b"my bytes", "text/plain"),
)

The async client uses the exact same interface. If you pass a PathLike instance, the file contents are read asynchronously automatically.

Handling errors

When the library is unable to connect to the API, or if the API returns a non-success status code (that is, 4xx or 5xx response), a subclass of APIError is raised:

import anthropic
# ...
try:
    message = client.messages.create(
        max_tokens=1024,
        messages=[
            {
                "role": "user",
                "content": "Hello, Claude",
            }
        ],
        model="claude-opus-5",
    )
except anthropic.APIConnectionError as e:
    print("The server could not be reached")
    print(e.__cause__)  # an underlying Exception, likely raised within httpx2
except anthropic.RateLimitError as e:
    print("A 429 status code was received; we should back off a bit.")
except anthropic.APIStatusError as e:
    print("Another non-200-range status code was received")
    print(e.status_code)
    print(e.response)

Error codes are as follows:

Request IDs

For more information on debugging requests, see Request ID.

All object responses in the SDK provide a _request_id property which is added from the request-id response header so that you can quickly log failing requests and report them back to Anthropic.

message = client.messages.create(
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude"}],
    model="claude-opus-5",
)
print(message._request_id)  # e.g., req_018EeWyXxfu5pfWkrYcMdjWG

Retries

Certain errors are automatically retried 2 times by default, with a short exponential backoff. Connection errors (for example, because of a network connectivity problem), 408 Request Timeout, 409 Conflict, 429 Rate Limit, and >=500 Internal errors are all retried by default.

You can use the max_retries option to configure or disable this:

# Configure the default for all requests:
client = Anthropic(
    max_retries=0,  # default is 2
)

# Or, configure per-request:
client.with_options(max_retries=5).messages.create(
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude"}],
    model="claude-opus-5",
)

Timeouts

By default requests time out after 10 minutes. You can configure this with a timeout option, which accepts a float or an httpx2.Timeout object:

import httpx2
from anthropic import Anthropic

# Configure the default for all requests:
client = Anthropic(
    timeout=20.0,  # 20 seconds (default is 10 minutes)
)

# More granular control:
client = Anthropic(
    timeout=httpx2.Timeout(60.0, read=5.0, write=10.0, connect=2.0),
)

# Override per-request:
client.with_options(timeout=5.0).messages.create(
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude"}],
    model="claude-opus-5",
)

On timeout, the SDK throws an APITimeoutError.

Note that requests that time out are retried twice by default.

Long requests

Avoid setting a large max_tokens value without using streaming. Some networks may drop idle connections after a certain period of time, which can cause the request to fail or timeout without receiving a response from Anthropic.

The SDK will throw a ValueError if a non-streaming request is expected to take longer than approximately 10 minutes. Passing stream=True or overriding the timeout option at the client or request level disables this error.

An expected request latency longer than the timeout for a non-streaming request will result in the client terminating the connection and retrying without receiving a response.

The SDK sets a TCP socket keep-alive option to reduce the impact of idle connection timeouts on some networks. This can be overridden by passing a custom http_client option to the client.

Auto-pagination

List methods in the Claude API are paginated. You can use the for syntax to iterate through items across all pages:

client = Anthropic()

all_batches = []
# Automatically fetches more pages as needed.
for batch in client.messages.batches.list(limit=20):
    all_batches.append(batch)
print(all_batches)

For async iteration:

async def main() -> None:
    all_batches = []
    async for batch in client.messages.batches.list(limit=20):
        all_batches.append(batch)
    print(all_batches)


asyncio.run(main())

Alternatively, you can use the .has_next_page(), .next_page_info(), or .get_next_page() methods for more granular control working with pages:

first_page = await client.messages.batches.list(limit=20)

if first_page.has_next_page():
    print(f"will fetch next page using these details: {first_page.next_page_info()}")
    next_page = await first_page.get_next_page()
    print(f"number of items we just fetched: {len(next_page.data)}")

# Remove `await` for non-async usage.

Or work directly with the returned data:

first_page = await client.messages.batches.list(limit=20)

print(f"next page cursor: {first_page.last_id}")
for batch in first_page.data:
    print(batch.id)

# Remove `await` for non-async usage.

Default headers

The SDK automatically sends the anthropic-version header set to 2023-06-01.

If you need to, you can override it by setting default headers on the client object or per-request.

# Set default headers for all requests on the client
client = Anthropic(
    default_headers={"anthropic-version": "My-Custom-Value"},
)

# Or override per-request
client.messages.with_raw_response.create(
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude"}],
    model="claude-opus-5",
    extra_headers={"anthropic-version": "My-Custom-Value"},
)

Type system

Request parameters

Nested request parameters are TypedDicts. Responses are Pydantic models which also have helper methods for things like serializing back into JSON (v1, v2).

Typed requests and responses provide autocomplete and documentation within your editor. If you'd like to see type errors in VS Code to help catch bugs earlier, set python.analysis.typeCheckingMode to basic.

Response models

To convert a Pydantic model to a dictionary, use the helper methods:

message = client.messages.create(...)

# Convert to JSON string
json_str = message.to_json()

# Convert to dictionary
data = message.to_dict()

Handling null vs missing fields

In responses, you can distinguish between fields that are explicitly null versus fields that were not returned (missing):

response = client.messages.create(
    model="claude-opus-5",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}],
)
if response.my_field is None:
    if "my_field" not in response.model_fields_set:
        print("field was not in the response")
    else:
        print("field was null")

Advanced usage

Accessing raw response data (for example, headers)

The "raw" Response returned by httpx2 can be accessed through the .with_raw_response property on the client. This is useful for accessing response headers or other metadata:

client = Anthropic()

response = client.messages.with_raw_response.create(
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude"}],
    model="claude-opus-5",
)

print(response.headers.get("request-id"))
message = (
    response.parse()
)  # get the object that `messages.create()` would have returned
print(message.content)

These methods return an APIResponse object. On the async client they return an AsyncAPIResponse, and .parse(), .read(), .text(), and .json() must be awaited.

Streaming response body

The .with_raw_response approach eagerly reads the full response body when you make the request. To stream the response body instead, use .with_streaming_response, which requires a context manager and only reads the response body once you call .read(), .text(), .json(), .iter_bytes(), .iter_text(), .iter_lines(), or .parse(). In the async client, these are async methods.

with client.messages.with_streaming_response.create(
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude"}],
    model="claude-opus-5",
) as response:
    print(response.headers.get("request-id"))

    for line in response.iter_lines():
        print(line)

The context manager is required so that the response will reliably be closed.

Logging

The SDK uses the standard library logging module.

You can enable logging by setting the environment variable ANTHROPIC_LOG to debug or info:

export ANTHROPIC_LOG=debug

Making custom/undocumented requests

This library is typed for convenient access to the documented API. If you need to access undocumented endpoints, params, or response properties, the library can still be used.

Undocumented endpoints

To make requests to undocumented endpoints, you can use client.get, client.post, and other HTTP verbs. Options on the client, such as retries, are respected when making these requests.

import httpx2

response = client.post(
    "/foo",
    cast_to=httpx2.Response,
    body={"my_param": True},
)

print(response.json())

Undocumented request params

If you want to explicitly send an extra parameter, you can do so with the extra_query, extra_body, and extra_headers request options.

Undocumented response properties

To access undocumented response properties, you can access the extra fields like response.unknown_prop. You can also get all extra fields on the Pydantic model as a dict with response.model_extra.

Configuring the HTTP client

The SDK sends requests with httpx2, an API-compatible fork of httpx. To customize the HTTP client, including proxies and transports, pass your own httpx2 client as http_client:

import httpx2
from anthropic import Anthropic, DefaultHttpxClient

client = Anthropic(
    # Or use the `ANTHROPIC_BASE_URL` env var
    base_url="http://my.test.server.example.com:8083",
    http_client=DefaultHttpxClient(
        proxy="http://my.test.proxy.example.com",
        transport=httpx2.HTTPTransport(local_address="0.0.0.0"),
    ),
)

You can also customize the client on a per-request basis by using with_options():

client.with_options(http_client=DefaultHttpxClient(...))

Tracing and mocking tools that patch httpx itself, such as OpenTelemetry's HTTPXClientInstrumentor, Sentry's httpx integration, respx, or pytest-httpx, do not see the SDK's requests by default. To use them, call httpx2.alias_httpx() once at startup, before anything imports httpx. This makes import httpx resolve to httpx2 for the whole process.

Managing HTTP resources

By default the library closes underlying HTTP connections whenever the client is garbage collected. You can manually close the client using the .close() method if desired, or with a context manager that closes when exiting.

with Anthropic() as client:
    message = client.messages.create(...)

# HTTP client is automatically closed

Beta features

Beta features are available before general release to get early feedback and test new functionality. You can check the availability of all of Claude's capabilities and tools in the build with Claude overview.

You can access most beta API features through the beta property of the client. To enable a particular beta feature, you need to add the appropriate beta header to the betas field when creating a message.

For example, to enable context editing:

client = Anthropic()

response = client.beta.messages.create(
    model="claude-opus-5",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude"}],
    betas=["context-management-2025-06-27"],
)

Platform integrations

All five client classes are included in the base anthropic package:

The AnthropicAWS client is in beta. Pass workspace_id to the constructor or set the ANTHROPIC_AWS_WORKSPACE_ID environment variable.

Use AnthropicBedrockMantle for new projects; AnthropicBedrock remains for existing applications using the Bedrock InvokeModel API.

Semantic versioning

This package generally follows SemVer conventions, though certain backward-incompatible changes may be released as minor versions:

1. Changes that only affect static types, without breaking runtime behavior.
2. Changes to library internals which are technically public but not intended or documented for external use.
3. Changes that aren't expected to impact the vast majority of users in practice.

Determining the installed version

If you've upgraded to the latest version but aren't seeing new features you were expecting, your Python environment is likely still using an older version. You can determine the version being used at runtime with:

print(anthropic.__version__)

Additional resources

• GitHub repository
• API reference
• Streaming Messages
• Tool use with Claude