"Opus 5 is a really bad model"

bbehnamoh about 3 hours ago 4 commentsRead Article on twitter.com

ZH version is available. Content is displayed in original English for accuracy.

⚡ Community Insights

Discussion Sentiment

50% Positive

Analyzed from 94 words in the discussion.

Discussion (4 Comments)Read Original on HackerNews

dgellow•1 minute ago

Could we replace the link with https://xcancel.com/i/article/2081697911847481502 ?

dev0p•12 minutes ago

It honestly feels bugged. Had to ditch it and went back to Sonnet 5. And the token usage is SO bad, especially on Openclaw: yesterday it somehow fully burned a whole Pro session usage in 3 minutes (!!!) to expose an already built service on tailscale. Just how??

K0balt•about 1 hour ago

My experience with opus 5 has surfaced many of the complaints listed it the tweet. It seems to be slightly brighter than 4.8, but the laziness more than makes up for it.

Bawoosette•about 2 hours ago

Most of this seems to be related to harness issues, specifically prompts provided by Claude Code. The upside of this is that it is likely cheap and easy for Anthropic to fix the issues by revising their approach to prompting their new models.