This is refreshing considering all the FUD (mostly from 1 frontier lab) happening around Open weight models.