I'm the maker of rrchat. Like many local AI enthusiasts, I love running open-weight LLMs (like Qwen, Llama, and DeepSeek) on my PC with llama.cpp. However, existing web interfaces are mostly built for desktops or require complex Docker configurations. Whenever I wanted to chat with my models from the couch or while on the go, the experience on mobile browsers was frustrating. I built rrchat to solve this exact problem: 1. Smartphone-first: Scan a QR code, add to Home Screen as a PWA, and you get a native mobile feel with safe-area support. 2. Access anywhere: Integrated with Tailscale Funnel, so you can reach your local PC without installing VPN apps on your phone. 3. Engine control: Directly orchestrates `llama-server`, tracks context and VRAM, and includes smart web search, reasoning blocks, and vision/OCR support. 4. Privacy-first & Shareable: Secured with personal tokens, plus isolated guest tokens if you want to let friends use your GPU without risking your PC. The project is 100% free and open-source I’d love to hear your thoughts, feedback, and what features you’d like to see next!

rrchat

A private AI chat for your local PC models via mobile phone

A private AI chat for your local PC models via mobile phone