Early testers claim Sonnet 5.5 outcodes rival flagships, sparking leaderboard drama

pranavmarla · x · 2026-09-28

chetaslua claims Sonnet 5.5 "mogs" rival flagship models at coding and Opus 5.5 beats others, sharing footage of Sonnet 5.5 in ultracode mode; the thread traces back to gauntlet-loop testing.

Community-level benchmark debate around the newest Anthropic models — treat specific claims as unverified.

Original post →

More from Models

Models channel →