OfflineLLM

OfflineLLM is a private on-device AI chat app for Android that runs any GGUF-format language model locally via llama.cpp, with ARM-optimized SIMD acceleration. It requires zero network permissions, encrypts all settings, and offers biometric lock and tamper detection to ensure data stays on your device. Designed for privacy-conscious users who want to experiment with open-weight models without sending data to the cloud. The app targets advanced users comfortable sourcing and managing their own model files, as there is no integrated model downloader. It supports only GGUF models and leverages ARM SIMD instructions for efficient inference on modern Android devices. Development appears to be a solo or small project with a changelog on GitHub. Who it's for: privacy enthusiasts, AI tinkerers, developers testing models on mobile, and anyone wanting offline AI chat without cloud dependencies. It is not for users expecting a turnkey experience or integration with hosted model services. What makes it different: complete network isolation, encryption and biometric security, and support for arbitrary GGUF models via llama.cpp — but limited documentation and no pricing page suggest it may be free or donation-supported.