Anthropic Engineer Addresses Claude Opus 5 Complaints
Anthropic engineers reply to reports that Opus 5 feels inconsistent and weaker than prior versions.
Users posting on X said Claude Opus 5 feels lazy, makes sloppy mistakes, and lacks the consistency of version 4.6. Engineer Thariq stated that no changes were made to the models and that the team is building easier ways to report regressions. He added that the team wants models to feel consistent and warm. Engineer Boris noted that verbosity has been a focus and that every model has different strengths. Kimmonismus said the feedback appears to have reached Anthropic and expressed hope for improvements.
Combined views
1.6M
12 posts, first seen 1d ago
Reactions from ranked influencers
12 posts
Wtf Anthropic: If Fable has felt noticeably dumber this week, it’s because Claude Code has apparently been running “high” reasoning effort at just 10/100, the same level that was „low“ before. They dumbed down fable without telling it seems.
if fable felt dumber this week, it's not you ❗❗❗ since 2.1.237 the model reads "high" effort as 10 out of 100, the exact number "low" used to be and the changelog doesn't say a word i spent my whole afternoon convinced t3 code and my own app were broken before i went digging…
Anthropic Thariq says Claude Code’s “high = 10” output was a numerical mapping issue, not a stealth downgrade. “High” is apparently still high, and internal evals show no performance regression. Thanks for the clarification. But I still have to say that most models currently…
We sometimes test API serving configs in Claude Code before rolling them out, and one running now maps the numerical effort value differently. That's why Claude may tell some of you it's at "10" on high. The scale isn't 0-100, the number isn't meaningful on its own, and the…
Thanks for amplifying! Nothing has changed on our end for the models but we’re working on rolling out something to make it much easier to report any feedback or regressions. I know it can be kind of hard to report feedback in the moment but it really helps us understand how…
You are welcome. I think I speak for many when I say again that Opus 5 unfortunately feels like a significant downgrade right now. It repeatedly feels lazy, makes sloppy mistakes, and if I catch it making a mistake, if it doesn't follow the instructions again, I constantly get…
@kimmonismus Thanks for the feedback. Yeah I agree that Opus 5 is a really spiky model and we want our models to be consistent and warm and feel like Claude. We’re working hard on this and it is a huge priority for us.
Thank you, Thariq! That's really good to hear. Lately, I've been a little worried that the feedback wasn't really getting through, because since 4.6, with 4.7 and 4.8, I felt like things were moving in a direction that went against community feedback. So it's all the better to…
