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: 总成本 ↑.
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） |
|---|---|---|---|---|---|---|---|
| #56 | GPT-5.4 Mini medium | OpenAI | 3 | 7.5 | $0.756 | 12/22 | 25.9s |
| #75 | Grok 4.20 medium | X AI | 2 | 7.1 | $0.777 | 12/22 | 29.5s |
| #99 | Qwen3.6 27B medium | Qwen | 1 | 6.5 | $0.779 | 10/22 | 106.3s |
| #73 | Grok 4.3 medium | X AI | 2 | 7.1 | $0.779 | 13/22 | 47.4s |
| #13 | GPT-5.3-Codex medium | OpenAI | 2 | 8.9 | $0.920 | 16/22 | 17.0s |
| #23 | Claude Sonnet 5 medium | Anthropic | 1 | 8.3 | $0.922 | 16/22 | 12.5s |
| #21 | GPT-5.2 medium | OpenAI | 3 | 8.4 | $0.951 | 14/22 | 22.6s |
| #31 | GLM 5.2 high | Z.ai | 1 | 8.0 | $0.970 | 14/22 | 62.7s |
| #28 | Inkling high | Thinkingmachines | 1 | 8.0 | $1.006 | 15/22 | 64.2s |
| #68 | Kimi K2.6 medium | Moonshot AI | 2 | 7.2 | $1.036 | 12/22 | 110.0s |
| #66 | Claude Opus 4.8 none | Anthropic | 1 | 7.3 | $1.166 | 13/22 | 4.91s |
| #16 | Muse Spark 1.1 medium | Meta | 2 | 8.6 | $1.357 | 15/22 | 25.0s |
| #18 | GPT-5.4 medium | OpenAI | 2 | 8.5 | $1.533 | 15/22 | 23.1s |
| #58 | Qwen3.5-27B medium | Qwen | 2 | 7.4 | $1.627 | 13/22 | 111.9s |
| #27 | Muse Spark 1.1 high | Meta | 2 | 8.1 | $1.694 | 12/22 | 31.5s |