orailnoor

orailnoor / termux-llm

Public

Run uncensored AI models locally on your Android phone. One-script setup via Termux + llama.cpp. No root, no cloud, no internet needed.

11
2
100% credibility
Found Apr 18, 2026 at 11 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Shell
AI Summary

This repository provides a one-script installer to run local, offline AI language models on Android devices via Termux.

How It Works

1
📱 Discover phone AI

You hear about a cool way to run your own smart AI helper right on your Android phone, completely private and offline.

2
🛒 Get Termux app

Download the free Termux app from F-Droid or the app store – it's like opening a little computer window on your phone.

3
🧠 Add AI brains

Make a folder named AI_Models in your phone's storage and drop in a downloaded AI model file, or let the setup grab one for you.

4
Run easy setup
🗂️
Pick your model

Choose from models you already saved if any are there.

📥
Grab a free one

Select a recommended uncensored model to download automatically.

5
🚀 Setup finishes

Sit back as it builds your AI engine – it takes 10-30 minutes the first time, then you're ready to go.

6
▶️ Start chatting

Type one quick command in Termux to launch, then open your phone's web browser and visit the local chat page.

💬 Your private AI

Enjoy talking to your smart, uncensored AI companion anytime – everything stays on your phone, no internet needed.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 11 to 11 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is termux-llm?

termux-llm lets you run uncensored local LLMs like Gemma-2-2B or Phi-3.5 on your Android phone via a single Shell script in Termux, powered by llama.cpp. Drop a GGUF model in /sdcard/AI_Models/, run the installer, and chat via browser at 127.0.0.1:8080 or hit the OpenAI-compatible API at /v1—no root, cloud, or internet needed after setup. It solves offline AI access on mobile, keeping everything on-device for privacy.

Why is it gaining traction?

One-command setup skips MLC LLM Termux hassles, auto-building llama.cpp and serving a web UI plus API that apps like SillyTavern can use. Uncensored "abliterated" models refuse nothing, and it auto-detects RAM to suggest fits, delivering 2-10 tokens/sec on mid-range phones. Devs dig the no-fuss path to termux local LLM without compiling from scratch.

Who should use this?

Android hackers wanting termux run LLM for offline coding assists or note-taking on the go. Privacy nuts ditching cloud LLMs for on-phone inference. Field devs in low-connectivity spots needing quick AI queries without data leaks.

Verdict

Grab it if you're experimenting with termux LLM on Android—solid docs and easy model swaps make it beginner-friendly despite 11 stars and 1.0% credibility score. Still early; test on non-critical setups as it's single-script simple but unproven at scale.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.