Secret Llama

Secret Llama is a privacy-first, open-source LLM chatbot that runs entirely in-browser, ensuring all conversation data remains local to your device.

Secret Llama is your secure, high-performance LLM client: a browser-based chatbot that processes all inference locally, guaranteeing zero data transmission to external servers. It supports popular open-source models like Llama and Mistral, leveraging WebGPU acceleration for responsive, in-browser execution. Once the model is cached, the tool functions completely offline, making it ideal for sensitive discussions or air-gapped environments. The Apache License V2.0 ensures transparency; your data stays on your machine, period.

https://github.com/abi/secret-llama