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） |
|---|---|---|---|---|---|---|
| #56 | Grok 4.20 Multi Agent Beta medium | X AI | 3 | 6.4 | 7/18 | 9.80s |
| #57 | GPT-5 Nano medium | OpenAI | 7 | 6.3 | 7/18 | 44.1s |
| #60 | Gemma 4 26B A4B none | 7 | 6.2 | 7/18 | 6.59s | |
| #62 | Gemini 2.5 Flash none | 10 | 6.2 | 7/18 | 903ms | |
| #63 | Qwen3.5-35B-A3B none | Qwen | 9 | 6.1 | 7/18 | 3.82s |
| #64 | DeepSeek V3.2 none | DeepSeek | 8 | 6.1 | 7/18 | 12.1s |
| #65 | MiMo-V2-Pro none | Xiaomi | 9 | 6.0 | 7/18 | 2.39s |
| #66 | GPT-5.4 none | OpenAI | 10 | 5.9 | 7/18 | 1.51s |
| #68 | gpt-oss-120b medium | OpenAI | 7 | 5.8 | 7/18 | 16.1s |
| #69 | Kimi K2.6 none | Moonshot AI | 8 | 5.8 | 7/18 | 2.05s |
| #50 | Hunter Alpha medium | OpenRouter | 4 | 6.7 | 8/18 | 10.3s |
| #54 | Mercury 2 medium | Inception | 6 | 6.5 | 8/18 | 2.21s |
| #55 | MiMo-V2-Omni none | Xiaomi | 8 | 6.5 | 8/18 | 1.99s |
| #58 | GLM 5V Turbo none | Z.ai | 8 | 6.2 | 8/18 | 3.10s |
| #59 | Qwen3.5-Flash none | Qwen | 9 | 6.2 | 8/18 | 3.25s |