setheerwagen/mcp-modelmanager

Overview

Two Model Context Protocol (MCP) servers for self-hosted, local AI. One manages and trains Ollama/vLLM models, the other routes questions and tasks to local LLMs through Claude, ChatGPT or local models. Runs on existing hardware and can access local data when wired in.

Best for

Best for
Developers building local-first AI workflows with MCP

Use cases

• Manage and train Ollama/vLLM models locally
• Route queries to local LLMs through MCP
• Let Claude or ChatGPT use local models and data

How to use

Install

pip install mcp-modelmanager

Tested with

ChatGPT

Notes

Two Model Context Protocol (MCP) servers for self-hosted, local AI. One manages and trains Ollama/vLLM models, the other routes questions and tasks to local LLMs through Claude, ChatGPT or local models. Runs on existing hardware and can access local data when wired in.

0 stars on GitHub. Last updated 2026-09-08. Licensed MIT.

Use cases

• Manage and train Ollama/vLLM models locally
• Route queries to local LLMs through MCP
• Let Claude or ChatGPT use local models and data

Pros

• Self-hosted, keeps data on your machines
• Uses existing hardware, no cloud compute required
• Written in Python, easy to extend

Cons

• No GitHub stars yet, early-stage project
• Requires manual wiring of MCP servers and local model setup
• Model management limited to Ollama and vLLM

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.