Secret Llama is a client-side LLM chatbot that runs entirely in the browser without any server infrastructure. It leverages the WebLLM project for WebGPU-based inference, allowing users to interact with models like Llama 3, Mistral, TinyLlama, and Phi1.5 directly in Chrome or Edge. All conversation data stays on the user’s machine, providing full privacy with no data transmission to external servers.