Run your first model with Ollama
Install Ollama, run a model, check whether it uses your GPU and learn the commands you will use every day.
Install Ollama
Download Ollama from the official site, ollama.com/download, and follow the guide for your system: Windows, macOS or Linux. Each guide lists the requirements from the official documentation.
Run a model
Open a terminal and run:
ollama run llama3.1The first time, Ollama downloads the model, about 4.9 GB for the 8B version. When the >>> prompt appears, type a question. Type /bye to leave.
Commands you will use every day
ollama listshows installed models.ollama psshows loaded models and whether they use the GPU or CPU.ollama stop MODELunloads a model from memory.ollama rm MODELdeletes a model to free disk space.
Good to know
Ollama listens on http://localhost:11434 and by default only accepts connections from your own computer. A model stays loaded for five minutes after the last request, then it is unloaded. The default context length is 4096 tokens.
For the full command list, see the Ollama cheat sheet. If something goes wrong, check the troubleshooting section of your system's guide.
Quick check
Which command shows whether a model runs on the GPU or CPU?
Quick check
How do you leave an Ollama chat?

