YZ Index
评测数据
当前展示：Run #340 | 2026-09-27 | 128 题库 | 公式 v7 | 判分 v6.4
数据公开说明：为防止题库污染和过拟合，题目原文和预期答案不公开。本页展示模型回答、得分、判分方式等透明数据。完整方法论请参阅方法论页面。
| 模型 | 代码执行 | 材料约束 | 工程判断 | 任务表达 | 诚信 | 主榜分 | 性价比 | 稳定性 | 可用性 | 逐题 |
|---|---|---|---|---|---|---|---|---|---|---|
| Grok 4 grok | 100.00 | 94.80 | 75.00 | 75.00 | 80.00 pass | 97.66 | 90.8 | 38.2 | 100.0 | |
| Claude Opus 4.7 claude | 97.00 | 94.80 | 100.00 | 50.00 | 90.00 pass | 96.01 | 29.7 | 47.7 | 100.0 | |
| GPT-o3 gpt | 97.00 | 94.80 | 100.00 | 75.00 | 90.00 pass | 96.01 | 51.3 | 80.0 | 95.0 | |
| Claude Sonnet 4.6 claude | 93.90 | 90.10 | 88.90 | 90.00 | 90.00 pass | 92.19 | 91.0 | 69.3 | 100.0 | |
| Qwen3 Max qwen | 79.60 | 94.80 | 30.60 | 45.00 | 80.00 pass | 86.44 | 99.3 | 27.0 | 100.0 | |
| 豆包 Pro doubao | 92.00 | 72.00 | 75.00 | 100.00 | 90.00 pass | 83.00 | 100.0 | 39.6 | 100.0 | |
| GPT-5.5 gpt | 72.00 | 87.10 | 100.00 | 100.00 | 90.00 pass | 78.80 | 76.1 | 53.0 | 100.0 | |
| Gemini 2.5 Pro gemini | 70.00 | 69.80 | 75.00 | 70.00 | 90.00 pass | 69.91 | 95.3 | 37.1 | 90.0 | |
| DeepSeek V4 Pro deepseek | 68.90 | 69.80 | 75.00 | 100.00 | 80.00 pass | 69.31 | 97.5 | 37.8 | 95.0 | |
| Gemini 3.1 Pro gemini | 67.00 | 69.80 | 100.00 | 75.00 | 90.00 pass | 68.26 | 84.7 | 29.5 | 100.0 | |
| GLM-4.6 zhipu | 67.00 | 44.80 | 75.00 | 75.00 | 80.00 pass | 57.01 | 94.9 | 71.7 | 75.0 |
API 访问：如需程序化访问评测数据，请使用我们的 API。