QuantaLLM

Run large language models entirely on your Android phone. Powered by llama.cpp with Hexagon NPU acceleration and ONNX Runtime — 100% offline, fully private inference on ARM64.