Claude goes down every day, even if only intermittently. I think the total downtime is reaching the level of hours. It would be good if they clarified whether this is an inability to scale under heavy demand or a side effect of GenAI assisted development.
Yesterday and today was both for Opus 4.6, their heaviest model. So I think it's load related.
Unfounded speculation but it's curious that it's soon after enabling 1M context tokens on the default plan.
I wonder if there are some long-tail load ramifications of this pricing change.