I made an app A.I.R.I, it runs LLMs locally on your phone. I’ve made a pretty big upgrade from its initial release and it’s starting to feel like something more than just a chat app.
The main idea now is: your phone = a personal AI server
It can: - run models locally - be accessed by other devices on your Wi-Fi - support voice conversations (TTS + STT) - handle documents with a simple RAG pipeline - manage and download models inside the app - keep chat history + user profiles for context - I also completely refactored the architecture so it’s modular and easier to extend (which was badly needed).
Still a work in progress, but this is the first time it feels like the original idea is actually working. Repo: Link
[link] [comments]