Will any AI model reach 1575 Math Arena Score by December 31, 2026?
As of September 5, 2026, the market gives “Will any AI model reach 1575 Math Arena Score by December 31, 2026?” a 68% chance of YES.
| YES odds | 68% |
|---|---|
| NO odds | 33% |
| Volume | $38,291 |
| Closes | January 1, 2027 |
Updated September 5, 2026 · Live data from Polymarket
About this market
**Frontier AI labs continue pushing math reasoning capabilities through iterative releases and specialized training, with Anthropic's Claude variants currently leading several math-focused leaderboards including LMArena math Elo ratings above 1530 and strong expected performance on MathArena's competition tracks.** OpenAI's GPT-5 series has hit saturation-level results like 100% on AIME 2025 subsets, while Google and others add incremental gains via models such as recent Gemini flashes. Harder MathArena components, including proof-based contests and Apex problems, remain below 75% for top systems, creating headroom. With roughly four months until year-end, upcoming model updates, agent scaffolding improvements, and potential new reasoning architectures from major labs represent the main catalysts that could shift whether any system crosses the targeted score threshold. Trader consensus reflects the rapid but uneven pace of progress across benchmarks.