Don't make Naked LLM calls. Protect your users and their data

Author here. I need to calculate the timing delay but I don't feel any difference in speed in my local chatbot app, where every prompt is filtered. I am on Mac M3, 16GB.