What is Ollama on Mac?
By Junaid, maker of Taskglass · Last reviewed
Runs AI language models locally on your Mac.
Ollama downloads and runs large language models on your own Mac. The Ollama app sits in the menu bar and the ollama server process loads models into memory when they're used. A loaded model can take many gigabytes of memory, roughly the size of the model, which is expected.
- Category
- Developer tools
- Made by
- Ollama
- Safe to quit:
- Yes
Shown in Activity Monitor as ollama, Ollama
Why is Ollama using so much CPU or memory?
- A model is loaded or generating text
- Large models that need lots of memory
- Several models loaded at once
What to do
- 1Use smaller models if memory is tight
- 2Unload models you aren't using ('ollama stop <model>' or wait for them to time out)
- 3Quit Ollama from its menu bar icon when not needed
Safe to quit? Yes.
Safe to quit; any running generation stops and models are unloaded.
Related processes
Part of the Taskglass Process Library