谜题求解：未遵循指令
谜题求解
未遵循指令
看看哪些 AI 模型在 谜题求解 上最容易遇到 未遵循指令，更快找出薄弱点。 排序方式: 响应时间（平均） ↓.
86/86
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） |
|---|---|---|---|---|---|---|---|
| #58 | Qwen3.5-27B medium | Qwen | 1 | 8.2 | $1.627 | 2/3 | 59.6s |
| #46 | DeepSeek V4 Pro high | DeepSeek | 1 | 6.9 | $0.200 | 1/3 | 56.8s |
| #143 | Gemini 3.1 Flash Lite high | 2 | 5.7 | $2.044 | 1/3 | 50.8s | |
| #163 | Gemini 3.1 Flash Lite Preview high | 1 | 7.7 | $2.310 | 2/3 | 46.7s | |
| #77 | Kimi K2.5 medium | Moonshot AI | 1 | 5.3 | $0.600 | 1/3 | 43.2s |
| #16 | Muse Spark 1.1 medium | Meta | 1 | 7.9 | $1.357 | 2/3 | 42.5s |
| #52 | Kimi K2.7 Code medium | Moonshot AI | 1 | 5.9 | $0.751 | 1/3 | 41.0s |
| #76 | DeepSeek V3.2 medium | DeepSeek | 1 | 7.0 | $0.078 | 1/3 | 37.7s |
| #31 | GLM 5.2 high | Z.ai | 1 | 6.0 | $0.970 | 1/3 | 33.7s |
| #179 | Ring-2.6-1T none | Inclusionai | 1 | 7.7 | $0.026 | 2/3 | 31.5s |
| #45 | DeepSeek V4 Flash high | DeepSeek | 1 | 8.2 | $0.042 | 2/3 | 26.1s |
| #68 | Kimi K2.6 medium | Moonshot AI | 1 | 6.0 | $1.036 | 1/3 | 25.1s |
| #172 | MiniMax M2.7 medium | Minimax | 1 | 5.9 | $0.163 | 1/3 | 24.9s |
| #171 | North Mini Code none | Cohere | 2 | 3.5 | $0.000 | 0/3 | 24.4s |
| #150 | DeepSeek V4 Flash none | DeepSeek | 1 | 3.1 | $0.044 | 0/3 | 23.7s |