I want to see Mythos scores
Datacurve@datacurve
Opus 4.8 is now on DeepSWE.
On the default high thinking effort, it scores 6% higher than Opus 4.7 xhigh, while also lowering average cost per task.
3:04 PM · May 31, 2026 · 15.4K Views
Theo of t3.gg argues SWE-Bench is unreliable as DeepSWE benchmark results show Opus 4.8 outperforming Opus 4.7
Opus 4.8 also reduced the average cost per task.
--0--
Quote posts
Sentiment
Sentiment unavailable for this story.
Cluster Engagement
-
Views
-
Comments
-
Reposts
-
Bookmarks
Expand data
Posts from X
Most Activity
Most ActivityTimeline
VIEWS90.5KBOOKMARKS132LIKES688RETWEETS23REPLIES39
Theo - t3.gg@theo
swe-bench is kind of a shitshow, and it makes evaluating LLMs hard. DeepSWE is the first agentic code bench that makes sense.
Datacurve@datacurve
Opus 4.8 is now on DeepSWE.
On the default high thinking effort, it scores 6% higher than Opus 4.7 xhigh, while also lowering average cost per task.
15hViews 90.5KLikes 688Bookmarks 132