Users say Claude Opus 5.5 got worse; others dispute a decline
Original title: Ask HN: Opus 5.5 suddenly dumb as rocks?
Commenters describe context rot and fast quota burn, while a cited daily benchmark is said not to show a drop.
A Hacker News thread asks whether Claude Opus 5.5 became much worse over about half a day, with worse output across agents and projects. Replies split: some cite context rot, forgotten instructions and quota burned in review loops, while one points to a public daily benchmark that does not show the model getting dumber. Users also compare Opus 5.5 with Fable and mention falling back to Opus 4.8.
Unconfirmed: Reports are anecdotal and disputed; no vendor confirmation of a model change.
- One user says long Opus 5.5 sessions forget prior instructions and scripts.
- Another says weekly Claude quota is exhausted within a day.
- A commenter links a daily Opus 5.5 benchmark that does not show decline.
- Comparisons include Fable 5.1 and fallback use of Opus 4.8.
From the original article
4 min read at news.ycombinator.com[](https://news. ycombinator. com/)Hacker Newsnew past comments ask show jobs submitlogin [](https://news.
By jahala · news.ycombinator.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.