LLM client library

quill is a small Python client that talks to several model providers through one API. This page is its quickstart.

Install #

pip install quill-llm

uv add quill-llm

poetry add quill-llm

Optional extras add a provider’s own SDK for streaming and batch jobs:

uv add "quill-llm[anthropic,openai]"

pip install "quill-llm[anthropic,openai]"

Configure #

quill reads its settings from the environment, or from a .env file in the working folder. Set the key for each provider you use.

Your first call #

from quill import Client

client = Client()
reply = client.chat("Explain recursion in one sentence.")
print(reply.text)

import asyncio
from quill import AsyncClient

async def main() -> None:
    client = AsyncClient()
    reply = await client.chat("Explain recursion in one sentence.")
    print(reply.text)

asyncio.run(main())

What the call sends, and what comes back:

Supported providers #

Pick a model by its name, such as client.chat(..., model="gpt-mini"). quill maps it to the provider’s API name.

Next #