| # | 模型 | 分数 |
|---|---|---|
| 1 | Gemini 4 Argon | 77.9% |
| 2 | DeepSeek V4.1-Flash 1017 | 77% |
| 3 | DeepSeek V4.1-Flash 0910 | 74.2% |
| 4 | Claude Opus 5 | 74% |
| 5 | MiniMax M3.1 Flash | 73.7% |
| 6 | GPT 5.6-Sol | 73% |
| 7 | Grok 4.7 | 71% |
| 8 | Kimi K3 | 67.5% |
| 9 | GLM 5.3 | 66.9% |
| 10 | DeepSeek V4-Pro 0813 | 62.7% |
| 11 | DeepSeek V4-Flash 0731 | 54.4% |