AI BENCHY 失败分析
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: 响应时间（平均） ↓.
| 排名 | 模型 | 公司 | 未遵循指令 次数 | 分数 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|
| #97 | Qwen3.5-9B medium | Qwen | 2 | 4.4 | 3/18 | 73.6s |
| #46 | Kimi K2.5 medium | Moonshot AI | 2 | 7.0 | 9/18 | 72.4s |
| #39 | Seed-2.0-Mini medium | Bytedance Seed | 1 | 7.5 | 11/18 | 69.7s |
| #11 | Gemini 3.1 Flash Lite Preview high | 1 | 8.4 | 12/16 | 68.8s | |
| #32 | Qwen3.5-Flash medium | Qwen | 1 | 7.8 | 11/18 | 66.7s |
| #10 | Qwen3.5-27B medium | Qwen | 2 | 8.4 | 13/18 | 53.0s |
| #27 | DeepSeek V3.2 medium | DeepSeek | 1 | 8.0 | 12/18 | 46.4s |
| #34 | Kimi K2.6 medium | Moonshot AI | 3 | 7.7 | 11/18 | 45.2s |
| #57 | GPT-5 Nano medium | OpenAI | 3 | 6.3 | 7/18 | 44.1s |
| #71 | MiniMax M2.5 medium | Minimax | 3 | 5.7 | 5/18 | 39.6s |
| #93 | GLM 4.7 Flash medium | Z.ai | 2 | 4.6 | 4/18 | 32.3s |
| #80 | MiniMax M2.7 medium | Minimax | 6 | 5.3 | 4/18 | 31.1s |
| #6 | Seed-2.0-Lite medium | Bytedance Seed | 2 | 8.6 | 13/18 | 30.4s |
| #30 | Step 3.5 Flash medium | Stepfun | 3 | 7.9 | 11/17 | 26.8s |
| #24 | Gemma 4 26B A4B medium | 1 | 8.0 | 13/18 | 25.0s |