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: 失败次数 ↑.
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） |
|---|---|---|---|---|---|---|---|
| #440 | Clef Omni default | Cloudflare | 1 | 0.2 | $0.011 | 2/12 | 880ms |
| #442 | Step 3.5 Flash none | Stepfun | 1 | 0.2 | $0.020 | 6/12 | 39.0s |
| #445 | LFM2-24B-A2B default | Liquid | 1 | 0.2 | $0.001 | 2/16 | 782ms |
| #450 | Nimble v3 9B default | Bespoke Labs | 1 | 0.8 | ~$0.001 | 1/7 | 363ms |
| #34 | Claude Sonnet 5.5 max | Anthropic | 2 | 0.9 | $8.557 | 18/23 | 84.3s |
| #38 | GPT-5.3-Codex medium | OpenAI | 2 | 0.9 | $1.318 | 17/23 | 18.9s |
| #40 | Muse Spark 1.3 high | Meta | 2 | 0.9 | $1.891 | 17/23 | 52.9s |
| #54 | Muse Spark Open low | Meta | 2 | 0.9 | $0.600 | 16/23 | 21.6s |
| #57 | Muse Spark 1.1 medium | Meta | 2 | 0.9 | $1.665 | 16/23 | 27.0s |
| #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 |
| #67 | Qwen3.8 27B high | Qwen | 2 | 0.9 | ~$0.298 | 17/23 | 166.3s |
| #75 | Claude Haiku 5.5 low | Anthropic | 2 | 0.9 | $0.061 | 16/23 | 7.51s |
| #82 | Nemotron 3.5 Super VL high | NVIDIA | 2 | 0.8 | $0.349 | 15/23 | 49.9s |
| #88 | Muse Spark 1.1 high | Meta | 2 | 0.8 | $2.570 | 13/23 | 37.4s |