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 supported6 Subscribers
Add a CodeTriage badge to pydantic-ai
Help out
- Issues
- Community Claude Agent SDK model integration (use Claude Code subscription with Pydantic AI): documentation or upstream scope?
- test: no direct test coverage for _griffe.py docstring handling (doc_descriptions / _infer_docstring_style)
- No unit tests for `_gateway_cost` gateway cost parsing
- Add residual `_gateway_cost` parsing and header test coverage
- `SummarizingCompaction` sends the summary request whole, so a summarizer with a smaller context window fails the compaction
- `ModelSettings.extra_body` is typed `object`: non-mappings type-check, and `None` crashes four OpenAI-derived models
- Pydantic AI was independently reviewed on Hlido — scored 78/100, right of reply open
- `GroqModel` replays gpt-oss reasoning as `<think>` text, which gpt-oss then copies into its replies
- Responses API (openai:) returns empty completions on a resumed deferred-tool run; Chat Completions (openai-chat:) completes
- Foundry partner-hosted DeepSeek Flash deployments silently clamp Responses API output at 4096 tokens regardless of max_output_tokens
- Docs
- Python not yet supported