What model do you use?

Thank you for the app! Looks great! I’m curious what models you use? Also, do you collect any data?

Hi Pavel,

Thanks for trying it out! We use 50+ models (like Qwen 3, Gemma 3, Phi 4 and more) available on Hugging Face, with all quantisation options - over 10k combinations total. During installation, we check your hardware and automatically pick the best setup for your computer, but you can always change it in settings.

Regarding data - we don’t collect any personal information. We only gather performance metrics and error reports to help improve the app, which you can turn off in settings. Everything runs locally, so you don’t need internet to use it (only for downloading models during onboarding).