Benchmark
HMMT Feb 26
HMMT February 2026 is a math competition benchmark based on problems from the Harvard-MIT Mathematics Tournament, testing advanced mathematical…
- Modality
- text
- Categories
- math, reasoning
- Openness
- unknown
- Source
- llm_stats
- Reported scores
- 12
Reported scores
llm_stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| DeepSeek-V4-Flash-0423 | DeepSeek | 0.919 | 2026-04-23 | evidence |
| DeepSeek-V4-Flash-Max | DeepSeek | 0.948 | 2026-04-23 | evidence |
| DeepSeek-V4-Pro-Max | DeepSeek | 0.952 | 2026-04-23 | evidence |
| GLM-5.1 | Z.ai | 0.826 | 2026-04-07 | evidence |
| GLM-5.2 | Z.ai | 0.925 | 2026-06-16 | evidence |
| Kimi K2.6 | Moonshot AI | 0.927 | 2026-04-20 | evidence |
| MAI-Thinking-1 | Microsoft | 0.849 | 2026-06-02 | evidence |
| Qwen3.6 Plus | Qwen | 0.878 | 2026-04-02 | evidence |
| Qwen3.6-27B | Qwen | 0.843 | 2026-04-21 | evidence |
| Qwen3.6-35B-A3B | Qwen | 0.836 | 2026-04-16 | evidence |
| Qwen3.7 Max | Qwen | 0.971 | 2026-05-19 | evidence |
| Qwen3.7-Plus | Qwen | 0.929 | 2026-05-31 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.