AI BENCHY 分类失败
谜题求解：答案错误
谜题求解
答案错误
看看哪些 AI 模型在 谜题求解 上最容易遇到 答案错误，更快找出薄弱点。 排序方式: 失败次数 ↑.
| 排名 | 模型 | 公司 | 答案错误 次数 | 分类得分 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|
| #7 | Gemini 3.5 Flash medium | 1 | 7.7 | 2/3 | 2.38s | |
| #24 | GPT-5.2 Chat none | OpenAI | 1 | 7.7 | 2/3 | 4.10s |
| #28 | Gemini 2.5 Flash medium | 1 | 7.7 | 2/3 | 3.18s | |
| #36 | Qwen3.5 Plus 2026-04-20 medium | Qwen | 1 | 8.2 | 2/3 | 17.7s |
| #38 | Grok 4.3 medium | X AI | 1 | 5.9 | 1/3 | 22.5s |
| #40 | Gemini 3.1 Flash Lite Preview medium | 1 | 7.7 | 2/3 | 5.30s | |
| #43 | MiMo-V2.5-Pro medium | Xiaomi | 1 | 6.7 | 1/3 | 5.31s |
| #44 | Gemini 3.1 Flash Lite medium | 1 | 7.6 | 2/3 | 1.95s | |
| #46 | Qwen3.6 35B A3B medium | Qwen | 1 | 8.0 | 2/3 | 5.95s |
| #47 | Grok Build 0.1 medium | X AI | 1 | 7.7 | 2/3 | 18.3s |
| #48 | Gemini 3 Flash Preview none | 1 | 7.7 | 2/3 | 1.05s | |
| #54 | GPT-5 Mini medium | OpenAI | 1 | 5.6 | 1/3 | 15.2s |
| #55 | GLM 5.1 medium | Z.ai | 1 | 8.2 | 2/3 | 31.6s |
| #59 | GLM 5V Turbo medium | Z.ai | 1 | 7.7 | 2/3 | 10.2s |
| #60 | Kimi K2.6 medium | Moonshot AI | 1 | 6.0 | 1/3 | 25.1s |