Built by the people behind Pydantic, with the same instinct: describe what you expect and let validation enforce it. Agent inputs and outputs are typed, dependencies are injected, and your editor knows the shape of everything. It feels like normal Python.
Reach for it when
- You already use Pydantic and want the same discipline for agents
- You want typed, validated output rather than parsing prose
- You prefer small and explicit over large and magic
Look elsewhere when
- You want a large ecosystem of prebuilt integrations
- Your team does not use type hints
What the lessons will cover
- 01Install and first agent
- 02Typed results
- 03Dependencies
- 04Tools
- 05Static and dynamic system prompts
- 06Streaming
- 07Testing agents
- 08A typed agent end to end
Others in typed output
Common questions
Is Pydantic AI free to use?
Pydantic AI 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 Pydantic AI?
Basic Python is enough. Functions, dictionaries and imports cover most of what Pydantic AI asks of you.
When should I not use Pydantic AI?
You want a large ecosystem of prebuilt integrations. Your team does not use type hints.
Lessons for Pydantic AI 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 →