Possibly due to moving compute from inference to training
replyMy purely unfounded, gut reaction to Opus 4.7 being released today was "Oh, that explains the recent 4.6 performance - they were spinning up inference on 4.7."
replyOf course, I have no information on how they manage the deployment of their models across their infra.
[dead]
replyI was there too, but honestly after today, 4.7 "feels" just as a bad. I was cynical, but also, kind of eager for the improvement. It's just not there. Compared to early Feb, I have to babysit EVERYTHING.
reply