YZ Index
评测数据
当前显示：Run #408 | 2026-11-02 | 154题题库 | 公式 v7 | 判定 v6.5
数据公开说明：为防止题库污染和过拟合，题目原文与期望答案不公开。本页面展示模型回答、评分、判定方式等透明数据。完整方法论请参见方法论页面。
| 模型 | 代码执行 | 资料约束 | 工学判断 | 任务表达 | 誠実性 | 主评分 | 性价比 | 安定性 | 可用性 | 详细操作 |
|---|---|---|---|---|---|---|---|---|---|---|
| Claude Opus 4.7 claude | 81.00 | 84.40 | 70.30 | 64.60 | 90.00 pass | 82.53 | 5.4 | 34.0 | 100.0 | |
| GPT-6 Astra gpt | 90.00 | 73.20 | 89.60 | 86.10 | 86.70 pass | 82.44 | 8.4 | 42.1 | 100.0 | |
| 豆包 Pro doubao | 88.70 | 68.40 | 88.00 | 90.20 | 88.30 pass | 79.57 | 95.8 | 42.1 | 98.9 | |
| GPT-6.1 Sol gpt | 83.70 | 73.90 | 87.80 | 88.80 | 86.70 pass | 79.29 | 38.8 | 37.7 | 100.0 | |
| Grok 4 grok | 75.40 | 83.50 | 48.70 | 76.70 | 76.70 pass | 79.05 | 24.5 | 34.5 | 100.0 | |
| GPT-6 Luna gpt | 82.60 | 74.30 | 81.50 | 85.60 | 83.30 pass | 78.87 | 100.0 | 38.6 | 100.0 | |
| Claude Sonnet 4.6 claude | 83.20 | 71.80 | 64.90 | 64.20 | 83.30 pass | 78.07 | 24.9 | 31.4 | 100.0 | |
| GPT-6 Sol gpt | 85.90 | 65.80 | 80.50 | 86.10 | 86.70 pass | 76.86 | 37.6 | 36.3 | 100.0 | |
| GPT-o3 gpt | 73.90 | 77.20 | 85.10 | 78.30 | 83.30 pass | 75.39 | 9.4 | 43.7 | 93.9 | |
| GPT-5.5 gpt | 78.40 | 69.10 | 90.00 | 90.80 | 86.70 pass | 74.22 | 18.6 | 36.1 | 97.0 | |
| DeepSeek V4 Pro deepseek | 66.60 | 82.30 | 81.70 | 88.20 | 71.70 pass | 73.67 | 47.1 | 49.7 | 89.7 | |
| Gemini 3.1 Pro gemini | 53.10 | 71.90 | 87.80 | 75.80 | 80.00 pass | 61.56 | 22.6 | 19.9 | 100.0 | |
| Gemini 2.5 Pro gemini | 39.10 | 66.50 | 69.90 | 62.70 | 81.70 pass | 51.43 | 29.1 | 29.5 | 79.8 | |
| GLM-4.6 zhipu | 19.90 | 52.10 | 22.80 | 41.30 | 48.30 warn | 34.39 | 20.7 | 41.4 | 41.9 | |
| Qwen3 Max qwen | - | - | - | - | - | - | 0.0 | 75.0 | 0.0 |
API访问：如需通过程序访问评测数据，请使用
API 。。