Run Ollama Models on Android

Ollama is a CLI-based tool that allows users to run and manage various AI models locally.

This module is designed for Android users who want to install and run Ollama on their devices using Termux. You need atleast 8GB RAM of android.

By following this guide, you will be able to utilize Ollama's capabilities directly from your Android device.

Termux is a terminal emulator that enables Android devices to run a Linux environment without requiring root access.

It is freely available and can be downloaded from the official Termux GitHub repository.

You need to set up termux first.

termux-setup-storage

This command will access your android storage.

pkg update && pkg upgrade

This will update your packages to new one. Press Y when needed.

pkg install git cmake golang

These packages provide Git for version control, CMake for software compilation, and Go, the programming language used to develop Ollama.

Clone Ollama's Github Repository

git clone --depth 1 https://github.com/ollama/ollama.git

Navigate to Ollama directory

cd ollama

Generate Go code

go generate ./...

Build Ollama

go build .

After that, start ollama server

./ollama serve &

Now ollama server will run, you can interact with it.

Now ollama already installed and running, you can run any ollama models.

You can check every models and library on ollama website.

For example, I will use deepseek-r1 models.

./ollama run deepseek-r1:8b

Last updated 5 months ago