llamafu pushes the boundaries of on-device LLM inference. Built with Flutter and llama.cpp, it runs full language models on mobile hardware with zero cloud dependency — studying the practical limits of memory, latency, and model quality on consumer devices.
Run full LLMs on Flutter mobile apps with zero cloud dependency — for privacy-preserving on-device AI.
llamafu is one option in a category that includes llama.rn, flutter_llama_cpp, Ollama Mobile , and private mobile LLM SDKs. Our Compare page has the full side-by-side.