Local AI Engineering with Ollama

Understanding What's Loaded

You can use the command ollama ps after running a model to see what's loaded in your system:

# Run the model to load it
echo "Hi there!" | ollama run granite3.3:2b

# Run the ps command
ollama ps

You should see something like this:

NAME             ID              SIZE      PROCESSOR    CONTEXT    UNTIL              
granite3.3:2b    07bd1f170855    1.9 GB    100% CPU     4096       4 minutes from now

Here's a quick explanation of the columns:

• NAME is the model tag you used to run it, in model:variant form. Same string you pass to ollama run or to the API's model field.