MathArena - HMMT Feb 2025: leaderboard
Metric: Accuracy (%). Source: matharena.ai. 60 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.2 (xHigh) | 100 |
| 2 | Best Expert | 100 |
| 3 | GPT-5.2 (High) | 98.33 |
| 4 | Step 3.5 Flash | 98.33 |
| 5 | Gemini 3 Pro (Preview) | 97.5 |
| 6 | GLM-5 | 97.5 |
| 7 | Gemini 3 Flash | 97.5 |
| 8 | DeepSeek V3.2 Speciale | 97.5 |
| 9 | GPT-5.5 | 96.5 |
| 10 | Claude Opus 4.8 | 96.5 |
| 11 | Grok 4 | 95 |
| 12 | Qwen 3.7 Max | 95 |
| 13 | Gemini 3.1 Pro (Preview) | 94.8 |
| 14 | DeepSeek V4 Pro | 94.4 |
| 15 | GLM-5.2 | 94.4 |
Interactive version: theaggregate.ai/benchmark?slug=matharena-hmmt-feb-2025 · How It Works · Data refreshed daily, snapshot 2026-09-05.