Orchestration · Project

Pydantic AI

Pydantic AI builds Python agents with typed tool inputs, dependencies and validated outputs.

+ Follow

Builder facts

  • Runtime. The package requires Python 3.11 or later. The slim install lets applications select provider and integration extras. Source ↗
  • Model access. Direct model calls use provider credentials. Pydantic AI Gateway is a separate route with its own key and spending controls. Source ↗
  • Structured output. An output type can be a Pydantic model, with the validated result returned alongside message history and usage. Source ↗
  • Dependency injection. Typed dependencies supply services and data to agent instructions, tools and output validators. Source ↗
  • Deferred tools. A run can return pending tool requests for approval or external execution. A follow-up run receives the original message history and the deferred results. Source ↗

Sources checked

Publisher update ·

Pydantic AI 2.71.0 ↗

Deferred tool requests expose validated arguments for approvers, and ModelRouter adds routing for agent runs.

Change timeline

0 stories

No verified releases in this timeline yet.