Which company has the best Text Arena Math AI model end of September?

Anthropic enters this market as a credible front-runner because it already looks strong in arena-style rankings and has recently been viewed by traders as the most likely company to hold the top math spot. That matters because the resolution source is not a broad academic benchmark but a specific leaderboard snapshot, which tends to reward the exact type of model family and release cadence that Anthropic has handled well in recent cycles. If the current pattern persists, Anthropic has a good chance of remaining at or near the top through the check date.

The main reason to avoid overconfidence is that the math-specific evidence is mixed. Some current math leaderboards put OpenAI models ahead, while other arena snapshots still show Anthropic in a leading position, which suggests the contest is tight rather than settled. The fact that different benchmarks point to different winners means there is real model-to-model variance, and a single new release or scoring update between now and September 30 could easily reshuffle first place.

The market’s yes price is higher than my independent estimate, which implies traders are leaning into Anthropic’s current strength and perhaps discounting the chance of a late surge by a rival. That makes sense because Anthropic does have the best combination of current visibility, strong arena performance, and a credible track record of competitive releases. Still, with several weeks left and other labs capable of rapid improvements, I would frame Anthropic as the favorite, not a lock, and assign a probability in the low 60s rather than the low 70s.