Recent leaderboard results on Text Arena math benchmarks show Google’s Gemini 4 series, including Argon and Carbon variants, holding the top spot with Elo-style scores around 1522–1544, while Anthropic’s Claude Opus 5 and Fable 5 models cluster immediately behind at 1507–1532. This narrow gap, combined with Anthropic’s consistent depth across multiple high-performing releases, underpins the 53.5% market-implied odds for Anthropic as second-best lab by end of December. Google’s 28.5% reflects its current lead but also the risk of slippage if Anthropic ships further refinements. Lower odds for OpenAI, Meta, and Chinese labs align with their trailing arena scores and slower recent math-specific gains. Traders are watching December model updates and any new benchmark votes that could shift these tight margins.
Polymarketデータを参照したAI生成の実験的な要約。これは取引アドバイスではなく、このマーケットの解決方法には一切関係ありません。 · 更新日1H6H1D7D30DALL
View resolved






















外部リンクに注意してください。
外部リンクに注意してください。
よくある質問