Instructor does one job well. You give it a Pydantic model, it gets the model to fill it in, and it retries on validation failure. No parsing prose, no stripping markdown fences, no regex over an answer.
Reach for it when
- You need reliable JSON out of a model
- You want retries on validation failure for free
- You want one small library, not a framework
Look elsewhere when
- You need an agent loop and tools
- Your provider already guarantees a schema
What the lessons will cover
- 01Install and setup
- 02Your first structured call
- 03Nested models
- 04Validation and retries
- 05Streaming partials
- 06Multiple providers
- 07Prompt tips
- 08Extraction at scale
Others in typed output
Common questions
Is Instructor free to use?
Instructor is open source and free to run yourself. You still pay whichever model provider you point it at, and this tutorial is free with no signup.
Do I need to know Python to use Instructor?
Basic Python is enough. Functions, dictionaries and imports cover most of what Instructor asks of you.
When should I not use Instructor?
You need an agent loop and tools. Your provider already guarantees a schema.
Lessons for Instructor are being written. Meanwhile the LangGraph tutorial covers the same ground: state, tools, loops, memory and human approval. Most of it carries straight over.
Start the LangGraph tutorial →