pydantic-rpc

Overview

PydanticRPC is a Python library that enables automatic generation of protobuf files from Pydantic model signatures, allowing developers to rapidly expose services via multiple protocols. The library provides comprehensive support for both gRPC and Connect-RPC, with built-in features like health checking, server reflection, and streaming capabilities. Its MCP support allows exposing services as AI assistant tools via both stdio and HTTP/SSE transports, making it versatile for AI integration scenarios.

Try asking AI

After installing, here are 4 things you can ask your AI assistant:

• ask(req: OlympicsQuery) -> CityLocation

  Query information about Olympic locations for a specific year

  Try:Where were the Olympics held in 2020?Try:Tell me about the 2004 Olympics location
• ask_stream(req: OlympicsDurationQuery) -> AsyncIterator[StreamingResult]

  Stream information about Olympic locations across a range of years

  Try:Show me Olympic locations from 2016 to 2020Try:Stream all Olympic venues between 2000 and 2010

Note: Tool names were inferred from the example code showing methods of the OlympicsAgent class. The actual MCP tool naming convention isn't explicitly documented in this README excerpt.

Comparable tools

Installation

pip install pydantic-rpc

pip install pydantic-rpc-cli

{
  "mcpServers": {
    "pydantic-rpc": {
      "command": "python",
      "args": ["-m", "pydantic_rpc.server"],
      "env": {
        "PYDANTIC_RPC_PACKAGE_NAME": "my.services"
      }
    }
  }
}

FAQ

What protocols does PydanticRPC support?

PydanticRPC supports gRPC, Connect-RPC, and MCP protocols, with built-in support for both stdio and HTTP/SSE transports for MCP.

Can I skip protobuf generation in production?

Yes, set the environment variable PYDANTIC_RPC_SKIP_GENERATION=true to skip runtime generation and use pre-generated modules.

Compare pydantic-rpc with