Pydantic AI
Pydantic is all you need.

The Pydantic Stack gives developers tools to build end-to-end type-safe applications, from local to prod, from AI to API.

Ship robust apps faster, in Python, TypeScript, Rust and others.

  • Schema [OSS]

    Pydantic Validation

    Maps the way

    Data validation for Python:

    • Most widely used data validation library for Python
    • Enforces standard type annotation
    • Plays nicely with type checkers, IDEs and your brain
    • Fast and extensible

  • Agents [OSS]

    Pydantic AI

    Speeds the way

    Agent framework / LLM library for Python:

    • Type safe and ready to build production grade applications
    • Model agnostic: Support for all popular LLM providers
    • OpenTelemetry integration

  • Observability [OSS + SAAS]

    Pydantic Logfire

    Lights the way

    OpenTelemetry traces, logs and metrics:

    • Developer tool and production monitoring
    • General purpose O11y with first-class support for AI
    • Custom SDK for Python, JS and Rust, standard OTel for other languages

Case Studies

All case studies
Sophos logo

Sophos's SecOps AI team implemented Pydantic Logfire for unified tracing across their AI-powered security solutions. With end-to-end visibility and SQL-based monitoring, engineers now detect issues proactively and run side-by-side LLM experiments with Pydantic Evals.

Read more→
BoostedAI logo

Boosted.ai implemented Pydantic Logfire for unified tracing and full-stack observability across 50,000+ AI research workflows. Allowing engineers to fnd and fix issues 12x faster, ensuring exceptional reliability and uptime for institutional finance clients.

Read more→
Mixam logo

Mixam, a global self-publishing company, partnered with Vstorm to build an AI agent using Pydantic AI that helps customers navigate complex printing specifications, reducing support burden while improving customer experience for all users who need ordering guidance.

Read more→

Complete AI Application Observability

Monitor LLM calls, agent reasoning, and your entire AI stack with Pydantic Logfire's AI observability platform — from local development to production.

Respected builders of open source, for Python & beyond

Pydantic is built on the pillars of open source: transparency, collaboration and excellence. Our origins inform every product we craft. Whether it’s a community project or enterprise solution, developer experience is our cornerstone.

Pydantic Validation is used and trusted by millions of individual developers - and some of the biggest organizations in the world, too.

We use Pydantic across key parts of our research and products, and it has accelerated our work considerably.

Atty Eleti Member of Technical Staff at OpenAI

Pydantic is incredibly powerful and a must use tool for AI engineers.

Logan Kilpatrick Lead product for Google AI Studio

We use Pydantic across key parts of our research and products, and it has accelerated our work considerably.

Atty Eleti Member of Technical Staff at OpenAI

Integrations that run on the Datadog Agent use Pydantic models that are generated from an OpenAPI spec to validate configuration which has drastically reduced customer confusion and support cases that we have to triage.

Ofek Lev Senior Software Engineer at Datadog

If you're not using Python yet, you should. If you're not using Pydantic yet with Python, you should. Pydantic is the data backbone of FastAPI, but even if you don't use FastAPI, Pydantic is extremely useful. There's always data, and handling data with Pydantic is several times more efficient and safer than without it and much more enjoyable.

Sebastián Ramírez creator of FastAPI
OUR MISSION

Driven by curiosity, frustration and coffee

Learn more about the creators of Pydantic, our mission and open source ethos. And see how you can get involved.

Our articles

/Pydantic Logfire

New timeline controls in Logfire

Laís Carvalho
Laís Carvalho
/Pydantic Logfire

How Tiger Data Built a Production AI Slack Bot with Pydantic AI and Logfire

Jacky Liang
Jacky Liang