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 ↓.
293/293
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） |
|---|---|---|---|---|---|---|---|
| #59 | Pareto 26.9 default | Unbiased | 1 | 0.9 | $1.864 | 15/23 | 28.3s |
| #60 | Qwen3.8 MAX low | Qwen | 1 | 0.9 | $1.089 | 17/23 | 25.1s |
| #61 | GPT-5.4 medium | OpenAI | 2 | 0.9 | $2.213 | 16/23 | 24.8s |
| #62 | Muse Spark Open medium | Meta | 2 | 0.9 | $1.260 | 15/23 | 77.4s |
| #63 | DeepSeek V4.1 Flash high | DeepSeek | 1 | 0.9 | $0.617 | 16/23 | 33.3s |
| #64 | DeepSeek V4.1 Flash medium | DeepSeek | 1 | 0.9 | $0.591 | 16/23 | 32.9s |
| #65 | Muse Spark 1.2 low | Meta | 3 | 0.9 | $0.881 | 16/23 | 11.8s |
| #66 | GPT-5.2 medium | OpenAI | 3 | 0.9 | $1.558 | 15/23 | 31.3s |
| #67 | Qwen3.8 27B high | Qwen | 2 | 0.9 | ~$0.298 | 17/23 | 166.3s |
| #68 | MiMo-V2.6-Pro medium | Xiaomi | 1 | 0.9 | $0.586 | 14/23 | 148.1s |
| #69 | Seed 2.1 Turbo medium | Bytedance Seed | 1 | 0.9 | $2.042 | 16/23 | 189.1s |
| #71 | DeepSeek V4.1 Flash max | DeepSeek | 1 | 0.9 | $1.204 | 15/23 | 61.6s |
| #72 | Muse Spark 1.1 low | Meta | 3 | 0.9 | $0.885 | 14/23 | 13.2s |
| #73 | Claude Sonnet 5.5 xhigh | Anthropic | 1 | 0.9 | $1.516 | 16/23 | 15.0s |
| #75 | Claude Haiku 5.5 low | Anthropic | 2 | 0.9 | $0.061 | 16/23 | 7.51s |