Skip to content

ollama CLI

ollama serve

  • You should run it to start the ollama server, it's necessary for all API operations (e.g., pull, run, etc)
  • You can also autostart it on login: brew services start ollama
ollama serve

ollama list

  • List all the downloaded models
ollama list

ollama ps

  • List running and exposed models
ollama ps

ollama pull

  • Download a model only
  • Models are saved at ~/.ollama/models/
  • After a model is pulled, it's ready to be used via API
ollama pull nomic-embed-text

ollama run

  • Download, run a model and plug it to stdin
  • If you want just to download -> ollama pull
ollama run gpt-oss
ollama run gemma3:270m

# same! (the model must be pulled first)
curl http://localhost:11434/api/chat \
  -d '{
        "model": "gpt-oss",
        "messages": [{"role": "user", "content": "Hello!"}]
      }'

ollama stop

ollama stop gpt-oss

ollama rm

ollama rm gpt-oss