HMMT February 2026: leaderboard
Official Hugging Face benchmark for model performance on the February 2026 Harvard-MIT Mathematics Tournament problem set.
Metric: Score (mixed sources). Source: huggingface.co. Source: benchmarklist.com. Source: huggingface.co. Saturation forecast: Around July 2027. 25 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.7 Max | 97.1 | #75 |
| 2 | GPT-5.5 | 96.7 | #31 |
| 3 | Claude Opus 4.8 | 96.7 | #37 |
| 4 | Claude Opus 4.6 (Max) | 96.2 | #63 (Claude Opus 4.6) |
| 5 | DeepSeek V4 Pro | 95.2 | #102 |
| 6 | Qwen 3.7 Plus | 92.9 | #92 |
| 7 | Kimi K2.6 (Thinking) | 92.7 | #104 (Kimi K2.6) |
| 8 | GLM-5.2 | 92.5 | #73 |
| 9 | Inkling Small | 90.2 | #138 |
| 10 | Qwen 3.5 397B A17B | 87.88 | #146 |
| 11 | Qwen 3.6 Plus | 87.8 | #120 |
| 12 | Gemini 3.1 Pro (Preview) | 87.3 | #58 |
| 13 | Kimi K2.5 | 87.12 | #144 |
| 14 | GLM-5 | 86.36 | #141 |
| 15 | Step 3.5 Flash | 86.36 | #270 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=hmmt-february-2026 · How It Works · Data refreshed daily, snapshot 2026-10-20.