Happy slice of mcp pizzaMCP.pizza
mcp-client-for-ollama

mcp-client-for-ollama

MCP.Pizza Chef: jonigl

This is a terminal-based chat application you install and run on your own computer, connecting free local AI models (through Ollama) to outside tools and information sources. You can browse files, run tools, and read documents the AI pulls in, while approving each action before it happens. It supports multiple tool sources at once, switching models mid-conversation, and saving your chat history. Setup requires installing Python and a package manager, and you can optionally add a paid key for cloud models like OpenAI, Claude, or Gemini.

Coding
Other

Use This MCP client To

Chat with a local AI model without sending my data online Let the AI use my files and tools safely Approve or block every tool action before it happens Switch between different local AI models mid-conversation Connect several tools to one conversation at once Save and revisit past conversations later

README

A simple yet powerful Python client for interacting with Model Context Protocol (MCP) servers using Ollama, allowing you to harness local LLMs for advanced tool execution.

English | 简体中文 | Español

MCP Client for Ollama (ollmcp)

PyPI - Downloads Python 3.11+ PyPI - Python Version PyPI - Python Version CI

Sponsored by
Atlas Cloud
Learn how to use Atlas Cloud with ollmcp in the Sponsors section

MCP Client for Ollama Demo

🎥 Watch this demo as an Asciinema recording

Table of Contents

mcp-client-for-ollama FAQ

Do I need to know how to code to use this?
You'll need to be comfortable with basic command line steps, such as installing Python and typing commands in a terminal window. It isn't a point-and-click app, so some technical comfort helps.
Does this cost anything or require a paid subscription?
No — it runs local AI models through Ollama on your own computer for free. You can optionally connect paid providers such as OpenAI, Claude, or Gemini instead, but that isn't required.
Do I need an API key?
Not if you stick with local Ollama models, which run entirely on your machine. If you add a cloud provider such as OpenAI, Claude, or Gemini, you'll need to supply your own key for that provider.
Can I use this to keep my conversations completely private?
Yes — when you use local Ollama models, your chats and files never leave your computer because everything runs offline.
Can I use this to safely test a new AI tool before trusting it?
Yes — its review-and-approve feature lets you see exactly what a tool wants to do and confirm it before it runs.
Which apps does this work with?
It's a standalone application you run in a terminal window rather than a plugin for an existing chat app. It connects to your local Ollama models and to any Model Context Protocol tool source you add.
How hard is it to set up?
It takes real setup — installing Python, the Ollama app, and a package manager, then running an install command. It's built for people comfortable with basic developer tools.
Can I switch AI models without restarting the app?
Yes — you can switch between any locally installed Ollama model right from within a running conversation.