MathArena - APEX 2025: leaderboard
Metric: Accuracy (%). Source: matharena.ai. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.8 (Max) | 81.25 |
| 2 | GPT-5.5 (xHigh) | 80.21 |
| 3 | Hy4 preview | 74.2 |
| 4 | GPT-5.4 Pro (xHigh) | 69.79 |
| 5 | Gemini 3.1 Pro (Preview) | 60.94 |
| 6 | GPT-5.4 (xHigh) | 54.17 |
| 7 | Qwen 3.7 Max | 44.5 |
| 8 | Claude Opus 4.7 (xHigh) | 40.62 |
| 9 | DeepSeek V4 Pro | 38.3 |
| 10 | Nex N2 Pro | 36.5 |
| 11 | Claude Opus 4.6 (Max) | 34.5 |
| 12 | Claude Opus 4.6 (High) | 34.45 |
| 13 | Gemini 3.5 Flash | 32.29 |
| 14 | DeepSeek V4 Pro (Max) | 28.12 |
| 15 | DeepSeek V4 Flash (Max) | 27.08 |
Interactive version: theaggregate.ai/benchmark?slug=matharena-apex-2025 · How It Works · Data refreshed daily, snapshot 2026-09-05.