Instructor

Combine LiteLLM with jxnl's instructor library for more robust structured outputs. Outputs are automatically validated into Pydantic types and validation errors are provided back to the model to increase the chance of a successful response in the retries.

Usage (Sync)

from litellm import completion
from pydantic import BaseModel

client = instructor.from_litellm(completion)

class User(BaseModel):
    name: str
    age: int

def extract_user(text: str):
    return client.chat.completions.create(
        model="gpt-4o-mini",
        response_model=User,
        messages=[
            {"role": "user", "content": text},
        ],
        max_retries=3,
    )

user = extract_user("Jason is 25 years old")

assert isinstance(user, User)
assert user.name == "Jason"
assert user.age == 25
print(f"{user=}")

Usage (Async)

from litellm import acompletion
from pydantic import BaseModel

client = instructor.from_litellm(acompletion)

class User(BaseModel):
    name: str
    age: int

async def extract(text: str) -> User:
    return await client.chat.completions.create(
        model="gpt-4o-mini",
        response_model=User,
        messages=[
            {"role": "user", "content": text},
        ],
        max_retries=3,
    )

user = asyncio.run(extract("Alice is 30 years old"))

assert isinstance(user, User)
assert user.name == "Alice"
assert user.age == 30
print(f"{user=}")