YZ Index
评测数据
当前展示:Run #409 | 2026-11-03 | 128 题库 | 公式 v7 | 判分 v6.5
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-5.5 gpt | 100.00 | 71.70 | 50.00 | 90.00 | 90.00 pass | 87.27 | 76.4 | 33.5 | 100.0 | |
| GPT-6 Astra gpt | 100.00 | 71.70 | 75.00 | 90.00 | 80.00 pass | 87.27 | 40.3 | 43.7 | 100.0 | |
| GPT-6 Luna gpt | 100.00 | 71.70 | 50.00 | 67.50 | 80.00 pass | 87.27 | 100.0 | 31.3 | 100.0 | |
| GPT-6 Sol gpt | 100.00 | 71.70 | 75.00 | 90.00 | 80.00 pass | 87.27 | 97.2 | 43.7 | 100.0 | |
| Claude Sonnet 4.6 claude | 88.90 | 80.00 | 50.00 | 32.50 | 80.00 pass | 84.90 | 86.7 | 26.0 | 100.0 | |
| GPT-6.1 Sol gpt | 100.00 | 59.20 | 75.00 | 90.00 | 80.00 pass | 81.64 | 96.6 | 42.4 | 100.0 | |
| DeepSeek V4 Pro deepseek | 75.00 | 80.00 | 75.00 | 90.00 | 100.00 pass | 77.25 | 99.0 | 49.0 | 95.0 | |
| Claude Opus 4.7 claude | 70.00 | 83.30 | 75.00 | 66.30 | 100.00 pass | 75.99 | 27.9 | 27.1 | 100.0 | |
| GPT-o3 gpt | 94.50 | 50.80 | 54.20 | 90.00 | 90.00 pass | 74.84 | 43.1 | 33.7 | 100.0 | |
| Gemini 3.1 Pro gemini | 96.90 | 46.70 | 75.00 | 60.00 | 90.00 pass | 74.31 | 84.9 | 25.0 | 100.0 | |
| 豆包 Pro doubao | 88.90 | 46.70 | 50.00 | 95.00 | 90.00 pass | 69.91 | 100.0 | 26.3 | 100.0 | |
| Grok 4 grok | 58.30 | 80.00 | 50.00 | 93.80 | 90.00 pass | 68.07 | 84.1 | 26.7 | 100.0 | |
| GLM-4.6 zhipu | 50.00 | 88.30 | 50.00 | 50.00 | 40.00 warn | 67.24 | 95.2 | 41.6 | 65.0 | |
| Gemini 2.5 Pro gemini | 70.80 | 55.00 | 60.50 | 83.80 | 90.00 pass | 63.69 | 96.1 | 35.7 | 95.0 | |
| 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。