Ollama MCP Bridge

by Community · Python

View source

Overview

Run local Llama, Mistral, and Qwen models via Ollama through MCP.

Set up

Runs locally

This server runs on your machine, so the install command depends on its package. The project README has the exact steps. Most clients expect an entry shaped like this:

mcp.json
{
  "mcpServers": {
    "ollama-mcp-bridge": {
      "command": "<command from the docs>",
      "args": [
        "<args>"
      ],
      "env": {}
    }
  }
}
Open README