Pydantic AI 2.57.0 released
A feature release adding Meta Muse model support, MCP tool improvements, and fixes for realtime streaming and token management.
What changed in 2.57.0
- New MetaProvider for Meta Muse models accessible via meta: prefix using the Responses API.
- MCP Apps tools can now be relayed to frontends and read resource contents directly.
- Harness ApprovalReviewer added to auto-approve low-risk tool calls before human review.
- Fixes for token-based history compaction, realtime transcript handling, and tool call metadata propagation across multiple model providers.
- Various fixes for streaming cleanup, concurrency management, and prompt caching behavior.
Summarised by AgentGid from Pydantic's official release notes on GitHub, as reported there; read the original before relying on it. Price, status and release dates come from our daily data.
Pydantic AI at a glance
| What it is | A Python agent framework from the Pydantic team, built around type-safe structured outputs, dependency injection and model-agnostic tool calling. |
|---|---|
| Latest version | 2.71.0, Nov 19, 2026. This page is about an earlier release. |
| Release pace | 13 stable releases in the last 30 days; 2.57.0 came 3 days after the previous one |
| Price | Free (OSS). Pydantic AI is open source and free to use; you pay only for the models and infrastructure it runs on. (checked Nov 4, 2026) |
| Downloads, last 7 days | 1.6M up 3.9% |
| Gid Score | 52 #15 in Agent frameworks and SDKs |