Meet Jarvis: The Open-Source French Voice Assistant Powered by Ollama and MCP A developer released jarvis-assistant-vocal, an open-source French-language voice assistant that runs fully offline on local LLMs via Ollama or connects to Claude for hybrid setups. The modular Python project combines native hardware tool execution (Hue, OBS) with Model Context Protocol server support and Twilio telephony, positioning itself as a more robust alternative to fragile DIY voice-agent scripts. jarvis-assistant-vocal delivers a production-ready, modular French voice assistant that runs entirely offline with local LLMs via Ollama, or connects to Claude for hybrid setups. By pairing native hardware tool execution with Model Context Protocol MCP server support, it cuts local automation complexity down to a single Python project. llama3 , mistral , etc. without rewriting logic. Ensure you have Python 3.10+ and your local Ollama instance running: 1. Clone the repository git clone https://github.com/sosoj92/jarvis-assistant-vocal.git cd jarvis-assistant-vocal 2. Set up virtual environment & install dependencies python -m venv venv source venv/bin/activate On Windows: venv\Scripts\activate pip install -r requirements.txt 3. Configure your environment cp .env.example .env Edit .env to set your LLM backend Ollama or Claude and peripheral API keys 4. Launch Jarvis python main.py Most DIY AI voice agents are fragile toy scripts that break the moment you hook them into real home automation or telephony. jarvis-assistant-vocal addresses this by adopting standard tool paradigms like MCP and combining hardware control Hue/OBS with telephony Twilio in one modular repository. If you are a self-hoster, smart home enthusiast, or streamer looking to run a private, self-hosted voice operator in French, this repo provides an exceptionally mature starting template. Building localized, private agents with small models? Follow Local AI & Infra Daily for the latest open-source breakthroughs, inference optimizations, and self-hosted infrastructure guides.