>>106402736
>does that chug resources?
llm is like 4GB and also available in smaller sizes. im using nodes to kill the ollama process when its done generating a prompt and that should also unload the model.

the fact that its a thinking model and also the reloading of the models chugs time