Harvard-MIT Mathematics Tournament February 2026 leaderboard
20 ranked models · higher is better
View accessible chart data
| Model | Rank | Provider | Score |
|---|---|---|---|
| Qwen3.7-Max | #1 | Alibaba Cloud | 97.1% |
| DeepSeek V4 Pro (Max) | #2 | DeepSeek | 95.2% |
| DeepSeek V4 Flash (Max) | #3 | DeepSeek | 94.8% |
| DeepSeek V4 Pro (High) | #4 | DeepSeek | 94% |
| Qwen3.7-Plus | #5 | Alibaba Cloud | 92.9% |
| Kimi K2.6 | #6 | Moonshot AI | 92.7% |
| GLM-5.2 | #7 | Z.AI | 92.5% |
| DeepSeek V4 Flash (High) | #8 | DeepSeek | 91.9% |
| Qwen3.5 397B | #9 | Alibaba Cloud | 87.9% |
| Qwen3.6 Plus | #10 | Alibaba Cloud | 87.8% |
| Kimi K2.5 | #11 | Moonshot AI | 87.1% |
| GLM-5 | #12 | Z.AI | 86.4% |
| Claude Opus 4.5 | #13 | Anthropic | 85.3% |
| MAI-Thinking-1 | #14 | Microsoft | 84.9% |
| Qwen3.6 27B | #15 | Alibaba Cloud | 84.3% |
| Qwen3.6-35B-A3B | #16 | Alibaba Cloud | 83.6% |
| GLM-5.1 | #17 | Z.AI | 82.6% |
| ZAYA1-8B | #18 | Zyphra | 71.6% |
| DeepSeek V4 Flash | #19 | DeepSeek | 40.8% |
| DeepSeek V4 Pro | #20 | DeepSeek | 31.7% |