---
id: instructor
version: "1.15.4"
license: MIT
license_treatment: permissive
maintenance: active
---

# instructor — structured outputs for llm

License: permissive · Maintenance: active · Downloads: 19.9M/mo

## What it is and what it does

Instructor is a Python library that bridges Pydantic models and LLM APIs to extract structured, validated data from natural language. Instead of writing JSON schemas, parsing responses, and handling validation errors manually, you define a Pydantic model and pass it to Instructor; the library generates the schema, sends it to the LLM, validates the response, and automatically retries on validation failure. It works with OpenAI, Anthropic, Google, Ollama, Groq, and other providers through a unified interface.

The package is built on Pydantic for type safety and IDE support, includes automatic retries with error feedback, supports streaming partial objects, handles nested data structures, and eliminates boilerplate around JSON parsing and manual validation. It's designed for developers who need reliable structured extraction without the complexity of writing custom parsing logic or managing provider-specific APIs.

Use it for:

- Extract product details (name, price, availability) from unstructured product descriptions or web scrapes.
- Parse customer feedback into structured fields (sentiment, category, action items) for downstream processing.
- Convert natural language requirements into typed configuration objects for application setup.
- Batch-extract entities (people, places, organizations) from documents with automatic validation and retry.
- Build chatbot intents and slot-filling workflows where LLM responses must conform to a known schema.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Instructor wraps LLM APIs to extract validated, typed structured data from natural language by defining Pydantic models and letting the package handle schema generation, validation, and retries.

Yes. Instructor is actively maintained, has no known vulnerabilities, low install friction, and a permissive license. It solves a real problem—reliable structured extraction from LLMs—with a clean API and broad provider support. Install it if you need to extract validated structured data from LLM responses; skip it if you're building agents or need richer observability (the docs suggest PydanticAI for that).

## Install

pip install instructor
uv add instructor
poetry add instructor

## Installing instructor

Before you install: Low friction: pure Python wheel with no compiled dependencies. Active maintenance—last commit 2026-08-09, 47 days since latest release. Eleven runtime dependencies (aiohttp, pydantic, openai, requests, tenacity, etc.) are all stable, widely-used libraries.

License in practice: MIT license is permissive; you can use, modify, and distribute this package freely in commercial and private projects without restriction.

Quickstart:

pip install instructor

import instructor
from pydantic import BaseModel

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

client = instructor.from_provider("openai/gpt-4o-mini")
user = client.chat.completions.create(
    response_model=User,
    messages=[{"role": "user", "content": "John is 25 years old"}],
)
print(user)

Requires Python 3.9 or later. You must have an API key for at least one supported LLM provider (OpenAI, Anthropic, Google, Ollama, Groq, etc.) and set it as an environment variable or pass it explicitly.

Verify before relying:

- Whether all 11 runtime dependencies are truly required for basic use or if some are optional for specific providers.
- Performance characteristics (latency, token efficiency) compared to raw LLM API calls or alternative extraction libraries.
- Coverage and reliability of validation retries across different LLM providers and edge cases.

## Package facts

- License: MIT (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 19.9M/month (top 5,000 on PyPI)
- Known vulnerabilities: none known

## Tags

structured output from llm, json extraction from language models, pydantic validation with llm, reliable llm structured data, llm response validation, extract typed objects from text, llm schema enforcement, llm-integration, data-validation, structured-extraction

[View on SkillFed](https://skillfed.io/packages/instructor) · [View on PyPI](https://pypi.org/project/instructor/)