AI BENCHY 失败分析
答案错误 失败
看看哪些 AI 模型最常遇到 答案错误，让你在选择前先发现稳定性风险。 排序方式: 测试正确 ↑.
| 排名 | 模型 | 公司 | 答案错误 次数 | 分数 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|
| #36 | GPT-5.3 Chat none | OpenAI | 5 | 7.7 | 11/18 | 5.88s |
| #38 | GPT-5.4 Nano medium | OpenAI | 4 | 7.6 | 11/18 | 11.2s |
| #39 | Seed-2.0-Mini medium | Bytedance Seed | 2 | 7.5 | 11/18 | 69.7s |
| #40 | GPT-5.2 medium | OpenAI | 2 | 7.5 | 11/18 | 14.0s |
| #41 | MiMo-V2-Flash medium | Xiaomi | 3 | 7.5 | 11/18 | 23.4s |
| #42 | Claude Sonnet 4.6 none | Anthropic | 3 | 7.4 | 11/18 | 4.98s |
| #30 | Step 3.5 Flash medium | Stepfun | 3 | 7.9 | 11/17 | 26.8s |
| #18 | GLM 5 Turbo medium | Z.ai | 3 | 8.1 | 12/18 | 17.7s |
| #23 | MiMo-V2-Pro medium | Xiaomi | 3 | 8.1 | 12/18 | 12.3s |
| #25 | Grok 4.20 Beta medium | X AI | 3 | 8.0 | 12/18 | 9.81s |
| #27 | DeepSeek V3.2 medium | DeepSeek | 3 | 8.0 | 12/18 | 46.4s |
| #28 | GPT-5.2 Chat none | OpenAI | 5 | 7.9 | 12/18 | 6.84s |
| #29 | Gemini 3.1 Flash Lite Preview none | 4 | 7.9 | 12/18 | 1.30s | |
| #33 | GLM 5.1 medium | Z.ai | 3 | 7.8 | 12/18 | 24.1s |
| #37 | Claude Opus 4.6 medium | Anthropic | 2 | 7.6 | 12/18 | 21.1s |