I've certainly noticed some variance from opus. there are times it gets stuck and loops on dumb stuff that would have been frustrating from sonnet 3.5, let alone something as good as opus 4.5 when it's locked in. But it's not obviously correlated with time, I've hit those snags at odd hours, and gotten great perf during peak times. It might just be somewhat variable, or a shitty context.
Now GPT4.1 was another story last year, I remember cooking at 4am pacific and feeling the whole thing slam to a halt as the US east coast came online.
>> ...or a shitty context
This is my guess, sometimes it churns through things without a care in the world and other times is seem to be intentionally annoying to eat up the token quota without doing anything productive.
Kind of have to see which mode it's in before turning it loose unsupervised and keep an eye on it just in case it decides to get stupid and/or lazy.