编程：答案错误
编程
答案错误
看看哪些 AI 模型在 编程 上最容易遇到 答案错误，更快找出薄弱点。
197/197
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 | 答案错误 次数 | 分类得分 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #207 | GLM 5V Turbo none | Z.ai | 2 | 5.5 | $0.052 | 1/3 | 3.13s |
| #209 | Gemma 4 26B A4B none | 2 | 3.7 | $0.015 | 0/3 | 4.16s | |
| #212 | KAT-Coder-Air V2.5 medium | Kwaipilot | 2 | 3.6 | $0.048 | 0/3 | 23.4s |
| #213 | Nemotron 3.5 Lightning high | NVIDIA | 2 | 4.4 | $0.134 | 0/3 | 168.4s |
| #214 | Qwen3.6 27B none | Qwen | 2 | 5.5 | $0.116 | 1/3 | 4.16s |
| #216 | Qwen3.8 27B none | Qwen | 2 | 5.5 | ~$0.006 | 1/3 | 1.19s |
| #222 | MiMo-V2.5-Pro none | Xiaomi | 2 | 4.3 | $0.068 | 0/3 | 1.41s |
| #223 | Kimi K2.5 none | Moonshot AI | 2 | 5.5 | $0.101 | 1/3 | 24.6s |
| #231 | Qwen3.6 35B A3B none | Qwen | 2 | 5.5 | $0.060 | 1/3 | 8.77s |
| #232 | Dots 3 Note Preview none | Dots Studio | 2 | 4.6 | $0.000 | 0/3 | 6.12s |
| #238 | MiMo-V2.5 none | Xiaomi | 2 | 5.5 | $0.025 | 1/3 | 3.24s |
| #244 | DeepSeek V3.2 none | DeepSeek | 2 | 3.1 | $0.054 | 0/3 | 14.5s |
| #262 | Qwen3 Coder Next medium | Qwen | 2 | 3.7 | $0.034 | 0/3 | 924ms |
| #274 | MiMo-V2-Flash none | Xiaomi | 2 | 4.3 | $0.025 | 0/3 | 2.64s |
| #5 | Gemini 3 Flash Preview medium | 1 | 8.6 | $0.763 | 2/3 | 84.4s |