Install and Use Ollama on Linux: A Beginner's Guide

What is Ollama?

Ollama is a tool for running large language models (LLMs) on your own computer. It downloads models, serves them, and lets you chat with them from the terminal.

Running a model locally means your prompts never leave your machine.

Install Ollama on Linux

The easiest way to install Ollama is with the official install script. Open a terminal and run:

curl -fsSL https://ollama.com/install.sh | sh

The script downloads Ollama and sets it up as a background service.

Check that the installation worked:

ollama --version

You should see output like ollama version 0.x.x.

Download a model

Ollama does not install models by default. You download one with ollama pull.

For a small, beginner-friendly model, use Llama 3.2:

ollama pull llama3.2

If your computer is low on memory, use the 1B version instead:

ollama pull llama3.2:1b

See which models you have downloaded:

ollama list

You should see llama3.2 listed with its size.

Load and interact with the model

The ollama run command loads a model and opens an interactive chat.

ollama run llama3.2

The model is now loaded. Type a prompt and press Enter:

Why is the sky blue?

The model streams its answer in the terminal. You can keep asking more questions.

You can also ask a single question without starting a chat:

ollama run llama3.2 "Tell me a short joke"

The model answers once and the command finishes.

Exit the interaction

To leave the interactive chat, type:

/bye

You can also press Ctrl+D.

You are now back at the shell prompt. The model is still installed and ready to run again whenever you need it.

What’s next?

Now that you can run a model, try other models with ollama pull. Remove models you no longer need with ollama rm. You can also use Ollama as a local API server for other tools and scripts.