编程 排名
看看哪些 AI 模型在 编程 上表现最好，哪些更稳定，以及差距主要出现在哪里。 排序方式: 测试正确 ↓.
289/289
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 | 编程 得分 | 分数 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #123 | Kimi K2.5 medium | Moonshot AI | 6.1 | 7.0 | $0.479 | 1/3 | 217.5s |
| #125 | Qwen3.7 Flash medium | Qwen | 6.2 | 6.9 | $0.065 | 1/3 | 46.8s |
| #126 | Seed-2.0-Code high | Bytedance Seed | 6.2 | 6.9 | $1.273 | 1/3 | 266.7s |
| #131 | Solar Pro 4 low | Upstage | 5.5 | 6.8 | $0.044 | 1/3 | 259.6s |
| #132 | Solar Pro 4 medium | Upstage | 5.5 | 6.8 | $0.047 | 1/3 | 310.4s |
| #133 | DeepSeek V4 Pro none | DeepSeek | 5.6 | 6.8 | $0.351 | 1/3 | 13.4s |
| #134 | GLM 5.3 low | Z.ai | 6.2 | 6.8 | $0.257 | 1/3 | 21.0s |
| #135 | MiMo-V2.5-Pro medium | Xiaomi | 6.2 | 6.8 | $0.221 | 1/3 | 92.1s |
| #136 | Gemini 3 Flash Preview none | 5.5 | 6.8 | $0.085 | 1/3 | 1.80s | |
| #139 | KAT-Coder-Pro V2.5 none | Kwaipilot | 6.1 | 6.7 | $0.487 | 1/3 | 22.5s |
| #140 | LongCat 2.0 low | Meituan | 6.6 | 6.7 | $0.412 | 1/3 | 479.3s |
| #141 | Gemini 3.5 Flash minimal | 5.6 | 6.7 | $0.300 | 1/3 | 2.75s | |
| #142 | GLM 5V Turbo medium | Z.ai | 6.0 | 6.7 | $0.457 | 1/3 | 63.4s |
| #146 | Qwen3.5-27B none | Qwen | 5.8 | 6.6 | $0.058 | 1/3 | 1.80s |
| #147 | Gemini 3.1 Flash Lite Preview low | 5.5 | 6.6 | $0.646 | 1/3 | 1.39s |