Gemma 4 actually running usable on an Android phone (not llama.cpp)
Reddit - Artificial Intelligence
1 min read
About this article
I wanted a real local assistant on my phone, not a demo. First tried the usual llama.cpp in Termux — Gemma 4 was 2–3 tok/s and the phone was on fire. Then I switched to Google’s LiteRT setup, got Gemma 4 running smoothly, and wired it into an agent stack running in Termux. Now one Android phone is: running the LLM locally automating its own apps via ADB staying offline if I want Happy to share details + code and hear what else you’d build on top of this. https://preview.redd.it/7vkbrlzfryvg1....
You've been blocked by network security.To continue, log in to your Reddit account or use your developer tokenIf you think you've been blocked by mistake, file a ticket below and we'll look into it.Log in File a ticket
Originally published on April 18, 2026.
Curated by AI News.