Protopia

LM studio is more than simply a war off chatting or exposing an OpenAI API - it also runs the inference engine which is a lot more complicated than just loading a compiled executable. (And the chat function probably also does context and different composition algorithms can have a big impact on quality). You also may well want tool calling - so that is another piece of functionality you need to check.

And then there is inference performance - there are always new performance enhancements, so you need to check whether your tool includes the most recent such tweaks.

Pressing an "Empty cache" button is NOT a fix to a 99% memory utilisation problem slowing down your computer - it is only a WORKAROUND. A fix would be too work out what was causing this and preventing it from happening in the first place.

My advice - check Advanced System Settings and ensure that the system is not setup to prioritise caching like a file server.