AI BENCHY 失败分析
API 错误 失败
看看哪些 AI 模型最常遇到 API 错误，让你在选择前先发现稳定性风险。 排序方式: 失败次数 ↑.
| 排名 | 模型 | 公司 | API 错误 次数 | 分数 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|
| #20 | Gemini 3.5 Flash none | 3 | 8.1 | 15/21 | 9.93s | |
| #33 | Hy3 preview medium | Tencent | 3 | 7.7 | 14/21 | 16.3s |
| #100 | Grok Build 0.1 none | X AI | 3 | 6.0 | 7/19 | 28.7s |
| #105 | Nemotron 3 Super medium | NVIDIA | 3 | 5.8 | 8/21 | 32.0s |
| #126 | gpt-oss-120b none | OpenAI | 3 | 5.4 | 6/19 | 21.6s |
| #136 | Elephant Alpha medium | Openrouter | 3 | 5.1 | 6/21 | 1.27s |
| #137 | Elephant Alpha none | Openrouter | 3 | 5.1 | 5/21 | 1.22s |
| #159 | Ling-2.6-1T none | Inclusionai | 3 | 4.3 | 3/21 | 7.72s |
| #35 | Gemini 3 PRO Preview medium | 4 | 7.6 | 14/21 | 9.05s | |
| #83 | Step 3.5 Flash none | Stepfun | 4 | 6.6 | 6/12 | 39.0s |
| #92 | Laguna M.1 medium | Poolside | 4 | 6.4 | 9/19 | 14.7s |
| #107 | Laguna Xs.2 medium | Poolside | 4 | 5.8 | 6/19 | 6.73s |
| #133 | DeepSeek V3.2 none | DeepSeek | 4 | 5.2 | 6/21 | 13.8s |
| #145 | Laguna M.1 none | Poolside | 4 | 4.8 | 4/19 | 2.89s |
| #146 | Laguna Xs.2 none | Poolside | 4 | 4.8 | 5/19 | 806ms |