LLM for WhatsApp is a free desktop app that turns WhatsApp into an AI-assisted inbox, automatically composing and sending replies powered by the large language model of your choice. The open-source project has more than 160 stars on GitHub and is released under the MIT license.
Instead of a cloud service, LLM for WhatsApp runs on your own computer — Windows, macOS or Linux. You download the release package, run the setup script for your platform, connect your WhatsApp account, and the app watches your incoming chats. When a message arrives in a chat where auto-reply is enabled, the app sends the conversation to your chosen model — OpenAI’s GPT models, a local model served by Ollama or llama.cpp, a Hugging Face endpoint, or another custom LLM API — and posts the generated reply for you. Conversation history is kept per chat so answers stay in context, and a customizable system prompt lets you define the assistant’s tone and role, for example a support agent for your small business. Setup is deliberately light: the release bundle installs what it needs and launches the app, with a manual path documented for developers. The project was actively updated into 2025.
LLM for WhatsApp is free and open source under the MIT license. You pay nothing for the app itself; if you connect a commercial API such as OpenAI you pay that provider’s usage fees, while local models run at no cost.
It is aimed at freelancers, small-business owners and community managers who answer a high volume of WhatsApp messages and want an AI first responder — especially those who prefer local models and keeping data on their own machine.