According to PolyBeats monitoring, Google's Gemini 4 Carbon currently ranks first on the AI model leaderboard with a score of 1538; Anthropic's highest-ranked model is Claude Opus 5.5 (High), released on September 22, with 1514 points.
The biggest variable this month is Anthropic's Claude Fable 5.2 and Claude Mythos 5.2, released on November 9, which Anthropic calls its "most capable" models.
On the prediction market Polymarket, the "Yes" probability for "Will Google have the strongest AI model on November 30?" is 87.5%. Two accounts with win rates above 80% in the AI leaderboard sector agree on whether Fable 5.2 can surpass Gemini 4 Carbon:
Account 1 has an 84% win rate across 25 markets traded in the AI leaderboard sector, with PnL +$1.5k and ROI +3.6%. This trade's position size is 1.2x their average position size in this sector.
Account 2 has a 90.5% win rate across 21 markets traded in the AI leaderboard sector, with PnL +$3.9k and ROI +3.1%. This trade's position size is 0.15x their average position size in this sector.
The market's resolution rules state that settlement is based on the overall ranking on the arena.ai text leaderboard as of 12:00 PM ET on November 30, using the version with style control turned off and filtered to "Models"; whichever company owns the top-ranked model is the one settled as the winner.
PS: OpenAI's "Yes" probability is only 1.15%, and its best model, GPT-5.6 Sol, ranks 23rd on the default leaderboard.
Subscribe to BlockBeats membership to view the full prediction market news content, unlocking complete account profiles, abnormal position-splitting operations, sector profit analysis, on-chain fund tracking, and exclusive related news content, keeping you one step ahead in global news insight.
To subscribe to BlockBeats membership, please add @PolyBeatsVip_Bot. After successful payment, the channel link will be sent automatically
---------------------------------
See the future sooner, follow @PolyBeats_Bot
See tomorrow, today. Follow @PolyBeatsEN
