11 min read
Running Gemma 4 Locally on iOS and Android: The Google AI Edge Gallery Guide
Step-by-step guide to run Gemma 4 locally on iPhone or Android via Google AI Edge Gallery. Learn how Multi-Token Prediction unlocks fast on-device agents.
Curated technical writing and architecture reflections categorized under #Gemma.
Step-by-step guide to run Gemma 4 locally on iPhone or Android via Google AI Edge Gallery. Learn how Multi-Token Prediction unlocks fast on-device agents.
Step-by-step guide to run Ollama in Termux on Android. Run lightweight Qwen 2.5, Gemma & Llama models with one-line install, WebUI, and no root.
Turn an old 2GB–4GB Android phone into an offline AI server using llama.cpp in Termux. Fix memory errors, optimize CPU threads, and prevent overheating.