编程：未遵循指令
编程
未遵循指令
看看哪些 AI 模型在 编程 上最容易遇到 未遵循指令，更快找出薄弱点。
38/38
筛选模型
没有模型匹配当前搜索和筛选条件。
| 排名 | 模型 | 公司 | 未遵循指令 次数 | 分类得分 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #177 | Claude Sonnet 5.5 low | Anthropic | 2 | 0.3 | $0.909 | 0/3 | 5.81s |
| #204 | Claude Sonnet 5.5 medium | Anthropic | 2 | 0.3 | $1.100 | 0/3 | 6.07s |
| #243 | Claude Fable 5.1 low | Anthropic | 2 | 0.3 | $3.395 | 0/3 | 6.75s |
| #56 | Qwen3.8 27B high | Qwen | 1 | 0.8 | ~$0.298 | 2/3 | 248.6s |
| #64 | Claude Haiku 5.5 low | Anthropic | 1 | 0.8 | $0.061 | 2/3 | 7.30s |
| #77 | DeepSeek V4 Flash 0731 high | DeepSeek | 1 | 0.6 | $0.576 | 1/3 | 252.7s |
| #107 | Gemini 3.5 Flash medium | 1 | 0.8 | $1.316 | 2/3 | 12.6s | |
| #108 | Qwen3.8 2.4T A95B high | Qwen | 1 | 0.6 | $2.949 | 1/3 | 160.5s |
| #130 | Mistral Large 4 high | Mistral | 1 | 0.3 | $1.391 | 0/3 | 1277.8s |
| #131 | Seed-2.0-Code high | Bytedance Seed | 1 | 0.6 | $1.448 | 1/3 | 266.7s |
| #139 | DeepSeek V4 Pro 0423 none | DeepSeek | 1 | 0.6 | $0.083 | 1/3 | 13.4s |
| #160 | Qwen3.8 Flash Next xhigh | Qwen | 1 | 0.7 | ~$0.195 | 2/3 | 192.2s |
| #178 | Claude Opus 4.8 none | Anthropic | 1 | 0.5 | $2.334 | 1/3 | 3.29s |
| #198 | GLM 5.2 none | Z.ai | 1 | 0.4 | $0.250 | 0/3 | 7.55s |
| #206 | Mercury 2.5 Preview high | Inception | 1 | 0.7 | $0.040 | 1/3 | 6.13s |