Reasoning on HMMT Feb. 2025
99.4ScoreGPT-5.2 (xhigh)
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5.2 (xhigh)2026.02 | 99.4 | |
| GLM-52026.02 | 97.9 | |
| Gemini-3.0 ProVariant=Pro2026.01 | 97.5 | |
| Gemini 3 Pro2026.02 | 97.3 | |
| GLM-4.72026.02 | 97.1 | |
| Kimi K2.52026.02 | 95.4 | |
| Audex 30B-A3BContext Length=1M, Tool-integrated reasoning=true2026.07 | 93.8 | |
| Nemotron-3-Puzzle-75B-A9BModel=Nemotron-3-Puzzle-75B-A9B, Precision=FP8, Tool Use=with tools2026.07 | 93.6 | |
| Claude Opus 4.52026.02 | 92.9 | |
| Nemotron-3-Puzzle-75B-A9BModel=Nemotron-3-Puzzle-75B-A9B, Precision=FP8, Tool Use=no tools2026.07 | 92.7 | |
| DeepSeek-V3.2 ThinkingThinking Mode=true2026.01 | 92.5 | |
| DeepSeek-V3.22026.02 | 92.5 | |
| Audex 30B-A3BContext Length=1M, Tool-integrated reasoning=false2026.07 | 92.2 | |
| Kimi-K2 ThinkingThinking Mode=true2026.01 | 89.4 | |
| GPT-5 HighVariant=High2026.01 | 88.3 | |
| MiMo-V2-FlashVariant=Flash2026.01 | 84.4 | |
| Claude Sonnet 4.5Variant=Sonnet 4.52026.01 | 79.2 | |
| Audex 2BContext Length=128K, Tool-integrated reasoning=false2026.07 | 68.5 | |
| Audex 2BContext Length=128K, Tool-integrated reasoning=true2026.07 | 67.3 | |
| Qwen3-Omni 30B-A3B ThinkingContext Length=64K2026.07 | 60.4 | |
| Qwen3.5-Omni Flash 35B-A3BContext Length=256K2026.07 | 59 | |
| Step-Audio R1.1 33BContext Length=64K2026.07 | 27 |