Pydantic AI 2.57 adds tool-approval review and MCP Apps routing
Updated Pydantic AI
Summary
The October 16 release adds ApprovalReviewer, MetaProvider and MCP Apps frontend integration, with realtime compatibility changes.
- Tool
- Pydantic AI
- Category
- Frameworks
- Date
- Open source
- Yes
- License
- MIT
- Pricing
- Repository licensing does not include hosted-service, model-provider or infrastructure costs. No service prices are claimed in this release note.
Pydantic AI released v2.57.0 on October 16, 2026.
What changed
- Harness ApprovalReviewer allows a model to review low-risk tool calls before they are shown to a person. That is a configurable review capability, not a guarantee that automated approval is appropriate for every action.
- MetaProvider adds Meta Muse model support through a meta-prefixed Responses API route.
- MCP Apps tool results can be forwarded to AG-UI and Vercel AI frontends. MCPToolset gains resource-content reading and app tool-call relay methods.
- Cost accounting prices requests at the service tier reported by the response; EvaluationReport gains a dataset name.
Compatibility checks
The maintainers explicitly flag changes to background-tool completion metadata, skill invocation flags and GPT-Live/Gemini Live session history, usage and reply waits. Teams using those paths should read the compatibility section, not treat this as a drop-in feature-only update. The changelog also fixes compaction, interrupted-stream spend accounting and cleanup behavior. These are maintainer reports, not measured reliability gains from our own tests.
License and cost
The release-tag license is MIT. Software licensing is separate from hosted-service, inference and infrastructure charges; this note does not claim a free hosted tier.
These changes are reported by the maintainers. We have not independently benchmarked the release.
Explore the Pydantic AI hub and frameworks category.
More from Pydantic AI
Pydantic AI 2.55.0 adds Conversation and prompt caching
Pydantic AIPydantic AI 2.55.0, released October 9, 2026, requires Python 3.11 and adds Conversation, a unified cache setting and Postgres stores.
Category: Frameworks
Pricing: Open source
Questions
What is Pydantic AI?
The product site describes it as a model-agnostic Python framework featuring type safety, structured outputs, validation and observability.
Is Pydantic AI open source?
The GitHub repository reports an MIT license.
Which models does it support?
The product site says it supports OpenAI, Anthropic, Google and more.
Filed on Agent Radar. Claims in this note are limited to the sources above.