YZ Index
评测数据
当前展示:Run #387 | 2026-10-20 | 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 |
|---|---|---|---|---|---|---|---|---|---|---|
| GPT-o3 gpt | 98.50 | 97.00 | 75.00 | 75.00 | 80.00 pass | 97.83 | 49.8 | 51.2 | 95.0 | |
| Claude Opus 4.7 claude | 97.00 | 84.10 | 75.00 | 53.90 | 90.00 pass | 91.20 | 27.9 | 34.6 | 100.0 | |
| GPT-5.5 gpt | 97.00 | 84.10 | 75.00 | 100.00 | 90.00 pass | 91.20 | 79.7 | 48.1 | 100.0 | |
| 豆包 Pro doubao | 95.70 | 84.10 | 75.00 | 100.00 | 80.00 pass | 90.48 | 100.0 | 47.1 | 100.0 | |
| GPT-6 Luna gpt | 97.00 | 79.30 | 61.70 | 100.00 | 100.00 pass | 89.04 | 100.0 | 47.6 | 100.0 | |
| Grok 4 grok | 92.80 | 84.10 | 75.00 | 75.00 | 55.00 warn | 88.89 | 87.7 | 27.2 | 100.0 | |
| Claude Sonnet 4.6 claude | 97.00 | 71.90 | 63.90 | 58.90 | 80.00 pass | 85.71 | 87.2 | 36.4 | 100.0 | |
| GPT-6.1 Sol gpt | 97.00 | 69.80 | 75.00 | 100.00 | 90.00 pass | 84.76 | 97.3 | 39.2 | 100.0 | |
| GPT-6 Sol gpt | 97.00 | 65.00 | 75.00 | 100.00 | 90.00 pass | 82.60 | 97.1 | 45.1 | 100.0 | |
| Gemini 2.5 Pro gemini | 75.00 | 70.00 | 50.00 | 85.00 | 90.00 pass | 72.75 | 95.6 | 25.6 | 95.0 | |
| GPT-6 Astra gpt | 72.00 | 69.80 | 75.00 | 100.00 | 90.00 pass | 71.01 | 35.8 | 29.0 | 100.0 | |
| Gemini 3.1 Pro gemini | 47.00 | 84.10 | 75.00 | 75.00 | 100.00 pass | 63.70 | 81.7 | 18.9 | 100.0 | |
| GLM-4.6 zhipu | 47.00 | 84.10 | 25.00 | 75.00 | 65.00 pass | 63.70 | 94.1 | 61.7 | 65.0 | |
| DeepSeek V4 Pro deepseek | 47.00 | 75.00 | 75.00 | 75.00 | 80.00 pass | 59.60 | 95.2 | 34.3 | 85.0 | |
| Qwen3 Max qwen | - | - | - | - | - | - | 0.0 | 75.0 | 0.0 |
API 访问:如需程序化访问评测数据,请使用我们的 API。