MCP Catalogs
Home

ollama-mcp-bridge

by jonigl·86·Score 47

Ollama MCP Bridge provides an API layer that integrates multiple MCP servers with the Ollama API, enabling dynamic AI tool integration for local LLM applications.

ai-llmdeveloper-toolsops-infra
27
Forks
11
Open issues
1 mo ago
Last commit
2d ago
Indexed

Overview

The Ollama MCP Bridge acts as a middleware service that sits between Ollama and multiple MCP servers, allowing Ollama to transparently use tools from all connected MCP servers. It provides a fully compatible API extension to the Ollama server, adding tool capabilities while maintaining compatibility with all existing Ollama endpoints. The bridge supports multiple transport types (stdio, HTTP, SSE), tool filtering, and configurable environments for different MCP servers.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

you:Building powerful local LLM applications with extended tool capabilities
you:Creating custom AI agents that can access multiple specialized tools
you:Enabling Ollama to use tools from different MCP servers simultaneously
you:Which Ollama endpoints are modified by the bridge?
you:How does tool filtering work?

When to choose this

Choose this when you want to extend local Ollama deployments with MCP tools without changing your existing application code.

When NOT to choose this

Not ideal if you need direct MCP integration without the Ollama layer or require write access to data sources (as most tools are read-only).

Tools this server exposes

4 tools extracted from the README
  • get_current_temperature

    Get the current temperature for a specified location

  • get_forecast

    Get a weather forecast for a specified location

  • read_file

    Read the contents of a file

  • list_files

    List files in a specified directory

Note: Tool names were inferred from the MCP server configuration example and tool filtering documentation. The actual tool implementations depend on the connected MCP servers.

Comparable tools

mcpnpxshell-mcp

Installation

Installation Options

**Quick Start with uvx:**

uvx ollama-mcp-bridge

**Install from PyPI:**

pip install --upgrade ollama-mcp-bridge

**Run with Docker:**

docker run -p 8000:8000 \
  -e OLLAMA_URL=http://host.docker.internal:11434 \
  -v "$PWD/mcp-config.json:/mcp-config.json" \
  ghcr.io/jonigl/ollama-mcp-bridge:latest

**Claude Desktop Configuration:** Add to Claude Desktop config.json:

"mcpServers": {
  "ollama-bridge": {
    "command": "ollama-mcp-bridge",
    "args": []
  }
}

FAQ

Which Ollama endpoints are modified by the bridge?
Only the `/api/chat` endpoint is modified to include MCP tools. All other endpoints are transparently proxied to the underlying Ollama server without modification.
How does tool filtering work?
You can configure include or exclude filters for each MCP server using the `toolFilter` configuration. Include mode allows only specified tools, while exclude mode blocks specified tools.

Compare ollama-mcp-bridge with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.