1824 points Claude Fable 5.2, ranked 1st in web development evaluation
Original headline1824점 클로드 페이블 5.2, 웹 개발 평가 1위
Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.
Anthropic’s Claude Fable 5.2 ranked first in Code Arena’s WebDev leaderboard with 1,824 points, six points ahead of Claude Opus 5.5 at 1,818 and above GPT-6.2 Astra at 1,797. The ranking, published on the 18th, covers 162 models and 1,027,059 total evaluation votes for web application-building performance. Fable 5.2 was tested under the xHigh setting and Opus 5.5 under Max, with reported score ranges of 1,824±17 and 1,818±11, respectively. Both Anthropic models had overlapping estimated rank ranges of first through third, leaving the top position statistically contested. The result matters because it signals tightening competition among frontier models on practical web development tasks. What to watch next is whether additional votes or configuration changes widen the gap or reshuffle the top three.
Muhabbit's AI wrote this summary in its own words from TokenPost's report and translated it from Korean. It can get details wrong, so check the source.
Named in this story
Claude Fable 5.2, Anthropic, Claude Opus 5.5, GPT-6.2 Astra, Code Arena