AI BENCHY 分类失败
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。 排序方式: 响应时间（平均） ↑.
134/134
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） |
|---|---|---|---|---|---|---|---|
| #141 | Laguna XS 2.1 none | Poolside | 3 | 4.3 | $0.003 | 0/3 | 623ms |
| #108 | Gemini 2.5 Flash none | 2 | 5.5 | $0.016 | 1/3 | 736ms | |
| #182 | Granite 4.1 8B none | IBM Granite | 1 | 4.5 | $0.003 | 0/3 | 775ms |
| #109 | Gemini 3.1 Flash Lite minimal | 2 | 5.5 | $0.013 | 1/3 | 831ms | |
| #112 | Qwen3.5-Flash none | Qwen | 2 | 5.5 | $0.005 | 1/3 | 850ms |
| #148 | Mistral Small 4 none | Mistral | 3 | 3.7 | $0.007 | 0/3 | 901ms |
| #143 | GPT-5.4 Mini none | OpenAI | 2 | 5.5 | $0.038 | 1/3 | 913ms |
| #167 | Qwen3 Coder Next medium | Qwen | 2 | 3.7 | $0.008 | 0/3 | 924ms |
| #111 | Gemini 3.1 Flash Lite none | 2 | 5.5 | $0.013 | 1/3 | 938ms | |
| #97 | Gemini 3.1 Flash Lite Preview none | 2 | 5.5 | $0.018 | 1/3 | 967ms | |
| #142 | GPT-5.6 Luna none | OpenAI | 3 | 3.8 | $0.047 | 0/3 | 980ms |
| #116 | GPT-5.6 Terra none | OpenAI | 2 | 5.5 | $0.130 | 1/3 | 1.00s |
| #114 | Nemotron 3 Ultra 550b A55b none | NVIDIA | 2 | 5.5 | $0.027 | 1/3 | 1.02s |
| #170 | Mercury 2 none | Inception | 3 | 3.4 | $0.011 | 0/3 | 1.03s |
| #168 | Grok 4.20 Beta none | X AI | 1 | 1.8 | $0.087 | 0/1 | 1.14s |