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 ↑.
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） |
|---|---|---|---|---|---|---|---|
| #402 | Grok 4.1 Fast medium | X AI | 4 | 0.4 | $0.069 | 9/19 | 23.8s |
| #290 | Gemini 3.1 Flash Lite none | 1 | 0.6 | $0.069 | 9/23 | 3.58s | |
| #372 | Qwen3 Coder Next medium | Qwen | 3 | 0.5 | $0.071 | 3/23 | 17.0s |
| #140 | Qwen3.7 Flash low | Qwen | 3 | 0.8 | $0.071 | 12/23 | 41.0s |
| #170 | Qwen3.8 27B medium | Qwen | 2 | 0.7 | ~$0.073 | 14/23 | 40.1s |
| #135 | GPT-6 Luna high | OpenAI | 1 | 0.8 | $0.074 | 15/23 | 20.4s |
| #221 | Qwen3.7 Flash medium | Qwen | 4 | 0.7 | $0.076 | 10/23 | 51.8s |
| #258 | Solar Pro 4 high | Upstage | 3 | 0.6 | $0.077 | 11/23 | 128.9s |
| #288 | Gemini 3.1 Flash Lite minimal | 3 | 0.6 | $0.080 | 10/23 | 3.95s | |
| #315 | Nemotron 3 Super medium | NVIDIA | 3 | 0.6 | $0.081 | 8/23 | 56.4s |
| #158 | DeepSeek V4 Pro 0423 none | DeepSeek | 2 | 0.7 | $0.083 | 10/23 | 14.1s |
| #411 | Grok 4.20 Beta default | X AI | 1 | 0.4 | $0.087 | 6/18 | 1.19s |
| #92 | Claude Haiku 5.5 high | Anthropic | 1 | 0.8 | $0.088 | 15/23 | 10.4s |
| #53 | GLM 5.3 Flash max | Z.ai | 1 | 0.9 | $0.089 | 17/23 | 54.0s |
| #145 | Qwen3.8 27B low | Qwen | 1 | 0.8 | ~$0.089 | 15/23 | 46.6s |