Anthropic

A Flyte task or app can call the Anthropic API with the anthropic Python SDK and a key from your own account.

Setting	Value
Python package	`anthropic`
API key	API keys
Environment variable	`ANTHROPIC_API_KEY`

Store your Anthropic API key

Run this command and paste the key at the prompt:

        
    
flyte create secret anthropic-api-key --project <your-project> --domain development

Call Anthropic from a task

call_anthropic.py
                
import os

import flyte

env = flyte.TaskEnvironment(
    name="anthropic-inference",
    image=flyte.Image.from_debian_base().with_pip_packages("anthropic"),
    secrets=flyte.Secret(key="anthropic-api-key", as_env_var="ANTHROPIC_API_KEY"),
)

@env.task
async def summarize(text: str) -> str:
    from anthropic import AsyncAnthropic

    client = AsyncAnthropic(api_key=os.environ["ANTHROPIC_API_KEY"])
    response = await client.messages.create(
        model="claude-opus-5-5",
        max_tokens=1000,
        messages=[{"role": "user", "content": f"Summarize in two sentences:\n\n{text}"}],
    )
    return response.content[0].text

if __name__ == "__main__":
    flyte.init_from_config()
    run = flyte.run(summarize, text="First document...")
    print(run.url)

Call Anthropic from an app

An AppEnvironment takes the same secrets argument, and the app reads the same environment variable. See Call the provider from an app.

Anthropic also has an OpenAI-compatible endpoint, which its docs describe as being for testing and comparing models, not production.

Route calls through the LLM Gateway

To add budgets, fallbacks, and guardrails, add Anthropic to the LLM Gateway using the same secret, then point your client at the gateway. See Choose direct calls or the gateway.