4 ms·

I know they (Ente) are working on local ML features for those types of use cases. Don't know the latest status.