models - anthropic - async

[pgrayy]

Description

Use the AsyncAnthropic Python client.

Related Issues

#83

Type of Change

• Bug fix
• New feature
• Breaking change
• Documentation update
• Other (please describe):

Testing

How have you tested the change? Verify that the changes do not break functionality or introduce warnings in consuming repositories: agents-docs, agents-tools, agents-cli

• I ran hatch run prepare
• Wrote new integ tests.
• Ran the following test script:

model = AnthropicModel(client_args={"api_key": "****"}, model_id="claude-3-7-sonnet-20250219", max_tokens=1000)
agent = Agent(model=model, callback_handler=None)

prompt = "What is 2+2? Think through the steps."


async def func_a():
    result = await agent.invoke_async(prompt)
    print(f"FUNC_A: {result.message}")


async def func_b():
    result = await agent.invoke_async(prompt)
    print(f"FUNC_B: {result.message}")


async def func_c():
    result = await agent.invoke_async(prompt)
    print(f"FUNC_C: {result.message}")


async def func_d():
    result = await agent.invoke_async(prompt)
    print(f"FUNC_D: {result.message}")


async def main():
    await asyncio.gather(func_a(), func_b(), func_c(), func_d())


asyncio.run(main())

Using the sync Anthropic Python client, every function ran sequentially and the total run time averaged 14s. Using the AsyncAnthropic Python client, every function ran concurrently and the total run time averaged 4s.

Checklist

• I have read the CONTRIBUTING document
• I have added any necessary tests that prove my fix is effective or my feature works
• I have updated the documentation accordingly
• I have added an appropriate example to the documentation to outline the feature, or no new docs are needed
• My changes generate no new warnings
• Any dependent changes have been merged and published

---

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[zastrowm]

Using the sync Anthropic Python client, every function ran sequentially and the total run time averaged 14s. Using the AsyncAnthropic Python client, every function ran concurrently and the total run time averaged 4s.

Wait, is this saying that wall time went from 14s -> 4s?!

[pgrayy]

Using the sync Anthropic Python client, every function ran sequentially and the total run time averaged 14s. Using the AsyncAnthropic Python client, every function ran concurrently and the total run time averaged 4s.

Wait, is this saying that wall time went from 14s -> 4s?!

Yes on my machine (very scientific lol). Of course this will all be hardware and prompt dependent. But no matter the case, customers can expect to see speed ups with concurrent executions.