pydantic-ai
https://github.com/pydantic/pydantic-ai
Python
GenAI Agent Framework, the Pydantic way
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported1 Subscribers
Add a CodeTriage badge to pydantic-ai
Help out
- Issues
- Output validator `ctx.last_attempt` can be inaccurate when `ToolOutput.max_retries` exceeds `output_retries`
- feat: implement strict runtime validation for tool outputs
- Expose agent as OpenAI Responses API endpoint via `Agent.to_responses()`
- Content filter error is not raised for `content_filter` finish reason
- fix: omit content key for tool-call-only assistant messages
- test: add tests for NativeOutput and PromptedOutput validation errors
- MCP isError:true is indiscriminately converted to ModelRetry — non-retryable tool errors should reach the LLM as informational results
- Ollama rejects assistant messages with content=None during multi-tool chaining
- Add `openai:gpt-5.5` and `openai:gpt-5.5-pro` to `KnownModelName`
- docs: fix indentation and typo in beta graph API warning admonition
- Docs
- Python not yet supported