LLM Hub
LLM Hub is an open-source mobile app for Android and iOS that runs chat, image, video and music generation entirely on the device, with CPU, GPU or NPU acceleration and a mode that builds a small app live from a spoken idea.
LLM Hub runs multiple on-device AI tools on a phone: multi-turn chat with retrieval-augmented memory, web search and multimodal input; an autonomous agent mode with function calling, device tools, MCP tools and, on Android, generated Termux shell commands; a custom-persona builder; image generation through Stable Diffusion 1.5; music and sound-effect generation through SoundGen on Android and Magenta Realtime 2 on iOS; image upscaling with AI super-resolution models; video generation through Stable Video Diffusion on iOS; offline translation and OCR across more than 50 languages; and speech-to-text transcription using Whisper models, all processed locally.
A vibe-coding mode explains an app idea in natural language and builds it with a live HTML and JavaScript preview on the device. The app supports multiple model formats so different models can run within the constraints of a phone's CPU, GPU or NPU.
Desktop versions for Windows and macOS are planned.