未遵循指令 失败
看看哪些 AI 模型最常遇到 未遵循指令，让你在选择前先发现稳定性风险。 排序方式: 分数 ↓.
140/140
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 | 分数 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #65 | Gemini 3.1 Flash Lite medium | 1 | 7.3 | $0.117 | 13/22 | 4.27s | |
| #66 | Claude Opus 4.8 none | Anthropic | 1 | 7.3 | $1.166 | 13/22 | 4.91s |
| #68 | Kimi K2.6 medium | Moonshot AI | 2 | 7.2 | $1.036 | 12/22 | 110.0s |
| #71 | Qwen3.7 Plus none | Qwen | 1 | 7.2 | $0.106 | 11/22 | 12.1s |
| #73 | Grok 4.3 medium | X AI | 2 | 7.1 | $0.779 | 13/22 | 47.4s |
| #75 | Grok 4.20 medium | X AI | 2 | 7.1 | $0.777 | 12/22 | 29.5s |
| #76 | DeepSeek V3.2 medium | DeepSeek | 1 | 7.0 | $0.078 | 11/22 | 68.6s |
| #77 | Kimi K2.5 medium | Moonshot AI | 2 | 7.0 | $0.600 | 10/22 | 99.0s |
| #78 | Mercury 2 medium | Inception | 3 | 7.0 | $0.093 | 10/22 | 2.72s |
| #80 | Seed-2.0-Mini medium | Bytedance Seed | 1 | 7.0 | $0.101 | 11/22 | 92.5s |
| #81 | KAT-Coder-Pro V2.5 medium | Kwaipilot | 1 | 6.9 | $0.467 | 11/22 | 24.0s |
| #82 | DeepSeek V4 Pro none | DeepSeek | 2 | 6.9 | $0.096 | 10/22 | 11.6s |
| #83 | GPT-5.6 Sol none | OpenAI | 1 | 6.9 | $0.524 | 11/22 | 2.16s |
| #84 | MiMo-V2.5-Pro medium | Xiaomi | 2 | 6.9 | $0.187 | 12/22 | 33.9s |
| #85 | Qwen3.6 Flash medium | Qwen | 1 | 6.9 | $0.738 | 12/22 | 44.7s |