🦙 Ollama MCP Server (Python)

Supercharge your AI assistant with local LLM access — run powerful AI models on your own computer, no internet required.

A Python MCP server that exposes your local Ollama models as tools for AI assistants like Windsurf, VS Code, Claude Desktop, and more.

Ecosystem

These two repositories work together:

┌─────────────────────────────────────┐
│  mcp-ollama-extension (VS Code)     │  ← start/stop, models UI, logs
│  https://github.com/pblagoje/       │
│       mcp-ollama-extension          │
└─────────────────┬───────────────────┘
                  │ stdio / MCP
┌─────────────────▼───────────────────┐
│  mcp-ollama-python (this repo)      │  ← MCP tools for Ollama
│  https://github.com/pblagoje/       │
│       mcp-ollama-python             │
└─────────────────┬───────────────────┘
                  │ HTTP :11434
┌─────────────────▼───────────────────┐
│  Ollama                             │
└─────────────────────────────────────┘

• This package is the MCP server (required). Install with pip install mcp-ollama-python.
• MCP Ollama Manager is the optional VS Code UI to manage that server. Prefer it if you use VS Code and want status bar controls, model sidebar, and log viewing without editing JSON by hand.

You can also wire this server into any MCP client (Windsurf, Cursor, Claude Desktop, etc.) without the extension — see Quick Start below.

What It Does

Connect your local LLMs to any MCP-compatible AI assistant. No cloud APIs needed.

Quick Start

Prerequisites: Python 3.10+, Ollama running locally

pip install mcp-ollama-python

Option A — VS Code with the companion extension (recommended for VS Code)

1. Install MCP Ollama Manager from the Marketplace
2. Ensure mcp-ollama-python is installed (the extension can offer Install Now if missing)
3. Command Palette → MCP Ollama: Configure Server → MCP Ollama: Start Server

Details: extension docs · repo

Option B — Any MCP client (Windsurf / VS Code MCP / Cursor / …)

Add to your MCP config (e.g. mcp_config.json):

{
  "mcpServers": {
    "ollama": {
      "command": "py",
      "args": ["-m", "mcp_ollama_python"],
      "disabled": false
    }
  }
}

Restart your editor — done. Your AI assistant can now use local Ollama models.

Try It

Type in your AI assistant's chat:

MCP Tool: ollama / ollama_chat — Use model llama3.1 and explain quantum computing

Key Features

• 🔧 8 MCP tools — Full Ollama SDK access
• 🔄 Hot-swap architecture — Drop a file in tools/, it's auto-discovered
• 🎯 Type-safe — Pydantic models throughout
• 🚀 Lightweight — Minimal dependencies, fast startup
• 🔌 Universal — Works with any MCP client

Documentation

Security (v1.0.8)

• ollama_execute is opt-in — set OLLAMA_EXECUTE_ENABLED=1; shell/bash removed
• OLLAMA_HOST validation — loopback by default; set OLLAMA_ALLOW_REMOTE_HOST=1 for LAN/remote Ollama
• No HTTP redirects — reduces SSRF risk toward internal endpoints

Details: docs/SECURITY.md

Changelog

1.0.8

• Security hardening: host validation, execute tool gating, env allowlist, input limits
• Fixed __version__ mismatch with package version
• Added tests/test_security.py and docs/SECURITY.md

Related Projects

License

MIT