Putnam 2025 (MathArena Accuracy)
Math · 2026-02-26
MathArena's Putnam 2025 evaluation across 12 problems. Scores are the leaderboard's Accuracy percentages; higher is better.
Top models (higher is better)
| Model | Score |
|---|---|
| DeepSeek-V3.2-Speciale | 85.8 |
| Gemini 3 Deep Think | 83.3 |
| GPT-5 Pro | 77.5 |
| Gemini 3 Pro Preview | 75.8 |
| GPT-5.1 | 48.3 |