73% used. 5 days remaining. I’m not making it to reset.

I’ve been getting some absurdly bad responses while using @claude with Opus 5 / Fable — bad enough to make me seriously question which model is actually answering. Looking deeper into my usage, I’m seeing charges for Opus 4.8 and Opus 4.7. So what are we actually getting? If I select Opus 5, I expect Opus 5 to answer — even if my question doesn’t require a “specialist.” I shouldn’t have to wonder whether my request was silently routed to a weaker model just because the system decided it was “good enough.” For basic questions, some of the answers have been shockingly poor. This lack of transparency is not acceptable. Very disappointing, @claude.

Camino a Miami ...