What it's used for

Instructor is a library that makes LLMs return structured, validated data by using Pydantic models as output schemas. It eliminates the unreliability of parsing free-text LLM outputs by leveraging function calling, tool use, or JSON mode to get typed Python objects directly.

Key use cases include:

Data extraction — pulling structured information from unstructured text (emails, documents, web pages)
API response formatting — ensuring LLM outputs match your application's data models
Classification — categorizing text with validated enum values
Entity extraction — extracting names, dates, amounts, and relationships from text
Validation with retries — automatically re-prompting when outputs fail Pydantic validation
Streaming structured data — partial object streaming for real-time extraction

Instructor is used by developers who need reliable structured outputs from LLMs in production. Created by Jason Liu, it has become the standard for structured extraction, supporting OpenAI, Anthropic, Google, Mistral, and more.

Available in Python, TypeScript (with Zod schemas), Ruby, Go, and Elixir.

Getting started

Install Instructor:
```
pip install instructor
```

Patch your LLM client and extract structured data:

import instructor
from openai import OpenAI
from pydantic import BaseModel

client = instructor.from_openai(OpenAI())

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

user = client.chat.completions.create(
    model='gpt-4o',
    messages=[{'role': 'user', 'content': 'John is 30, email john@example.com'}],
    response_model=User
)
print(user)  # User(name='John', age=30, email='john@example.com')