HMMT Feb 2026
5 models tested · Updated 2026-06-13 · Verified sources only
Qwen3.7-Max leads at 97.1%
1
Alibaba · Blog/Z.ai · 2026-06-13
Highest among compared models. Above Opus 4.8 (96.7).
97.1%
2
Anthropic · Blog/Z.ai · 2026-06-13
Ties GPT-5.5. Below DeepSeek V4 Pro (95.2)? No, above.
96.7%
3
DeepSeek · DeepSeek/HuggingFace · 2026-04-24
Mislabeled as AIME in DB; actually HMMT 2026 Feb (Pass@1) from official DeepSeek V4 Pro model card.
95.2%
4
Z.ai · Blog/Z.ai · 2026-06-13
Up from GLM-5.1 (82.6). 10-point gain.
92.5%
5
MiniMax · Blog/Z.ai · 2026-06-13
Well below frontier (Opus 4.8 96.7).
84.4%