AI BENCHY 分类失败
编程：超时
编程
超时
看看哪些 AI 模型在 编程 上最容易遇到 超时，更快找出薄弱点。 排序方式: 失败次数 ↑.
22/22
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 | 超时 次数 | 分类得分 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #30 | Qwen3.7 Plus medium | Qwen | 1 | 6.1 | $0.177 | 1/3 | 108.6s |
| #44 | Kimi K2.6 medium | Moonshot AI | 1 | 5.7 | $0.888 | 1/3 | 214.4s |
| #45 | Qwen3.5-122B-A10B medium | Qwen | 1 | 6.0 | $0.588 | 1/3 | 114.5s |
| #52 | MiniMax M3 medium | Minimax | 1 | 6.1 | $0.131 | 1/3 | 144.7s |
| #53 | DeepSeek V4 Pro high | DeepSeek | 1 | 6.1 | $0.157 | 1/3 | 243.0s |
| #56 | Kimi K2.5 medium | Moonshot AI | 1 | 6.1 | $0.348 | 1/3 | 217.5s |
| #61 | DeepSeek V3.2 medium | DeepSeek | 1 | 6.0 | $0.042 | 1/3 | 248.7s |
| #63 | Seed-2.0-Mini medium | Bytedance Seed | 1 | 5.5 | $0.044 | 1/3 | 220.5s |
| #72 | Gemma 4 26B A4B medium | 1 | 2.9 | $0.045 | 0/3 | 272.5s | |
| #75 | MiMo-V2-Flash medium | Xiaomi | 1 | 6.0 | $0.043 | 1/3 | 10.7s |
| #77 | GLM 5.1 medium | Z.ai | 1 | 4.6 | $0.288 | 0/3 | 109.6s |
| #79 | Kimi K2.7 Code medium | Moonshot AI | 1 | 7.6 | $0.581 | 2/3 | 146.7s |
| #84 | Qwen3.5-Flash medium | Qwen | 1 | 3.7 | $0.080 | 0/3 | 58.9s |
| #94 | Step 3.5 Flash medium | Stepfun | 1 | 2.4 | $0.070 | 0/2 | 258.4s |
| #104 | Qwen3.5-35B-A3B medium | Qwen | 1 | 5.9 | $0.401 | 1/3 | 206.6s |