Claude says Opus 5 is now the new SOTA on coding and knowledge-work benchmarks

chu_onthis · x · 2026-07-25

A repost citing Claude AI says Opus 5 is now the new state of the art on several coding and knowledge-work evaluations. The post itself adds only a brief “we cooked again” reaction, but the substantive claim is the benchmark result.

Related event: Anthropic Launches Claude Opus 5 with Leading Benchmark Performance(75 posts)→

Original post →

More from Models

Models channel →