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: 响应时间（平均） ↓.
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） |
|---|---|---|---|---|---|---|---|
| #68 | Qwen3.8 27B medium | Qwen | 10.0 | 7.8 | ~$0.058 | 3/3 | 40.5s |
| #12 | Gemini 3.1 Pro Preview medium | 7.9 | 9.2 | $1.352 | 2/3 | 40.2s | |
| #236 | Mistral Small 4 medium | Mistral | 4.4 | 5.1 | $0.097 | 0/3 | 40.0s |
| #206 | KAT-Coder-Air V2.5 high | Kwaipilot | 4.3 | 5.6 | $0.077 | 0/3 | 39.1s |
| #171 | gpt-oss-120b medium | OpenAI | 5.9 | 6.1 | $0.019 | 1/3 | 38.4s |
| #286 | Nemotron 3 Nano Omni 30b A3b Reasoning medium | NVIDIA | 1.1 | 3.4 | $0.000 | 0/1 | 38.1s |
| #211 | Owl Alpha none | Openrouter | 5.6 | 5.6 | $0.000 | 1/3 | 36.9s |
| #259 | Laguna M.1 medium | Poolside | 1.5 | 4.7 | $0.033 | 0/1 | 35.6s |
| #17 | Qwen3.7 Max medium | Qwen | 10.0 | 9.1 | $1.157 | 3/3 | 35.3s |
| #127 | Gemini 3.5 Flash none | 8.8 | 6.9 | $1.093 | 2/3 | 34.7s | |
| #110 | Muse Glimmer 30B medium | Meta | 8.4 | 7.2 | $0.280 | 2/3 | 33.7s |
| #67 | Claude Sonnet 4.6 medium | Anthropic | 5.7 | 7.8 | $2.126 | 1/3 | 33.3s |
| #124 | KAT-Coder-Pro V2.5 medium | Kwaipilot | 7.8 | 6.9 | $0.478 | 2/3 | 33.1s |
| #154 | Hy3 preview medium | Tencent | 5.3 | 6.5 | $0.049 | 1/3 | 31.4s |
| #183 | Grok 4.20 Beta medium | X AI | 3.3 | 6.0 | $0.750 | 1/1 | 31.4s |