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） |
|---|---|---|---|---|---|---|
| #15 | Gemini 2.5 Flash medium | 1 | 7.7 | 2/3 | 3.94s | |
| #17 | Gemini 3.1 Flash Lite Preview medium | 1 | 7.7 | 2/3 | 3.58s | |
| #21 | Gemini 3 Flash Preview none | 1 | 7.7 | 2/3 | 1.06s | |
| #27 | DeepSeek V3.2 medium | DeepSeek | 1 | 8.2 | 2/3 | 36.9s |
| #28 | GPT-5.2 Chat none | OpenAI | 1 | 7.7 | 2/3 | 4.42s |
| #30 | Step 3.5 Flash medium | Stepfun | 1 | 5.3 | 1/3 | 7.72s |
| #33 | GLM 5.1 medium | Z.ai | 1 | 8.2 | 2/3 | 23.8s |
| #34 | Kimi K2.6 medium | Moonshot AI | 1 | 5.0 | 0/3 | 25.6s |
| #35 | MiMo-V2-Omni medium | Xiaomi | 1 | 6.5 | 1/3 | 3.88s |
| #37 | Claude Opus 4.6 medium | Anthropic | 1 | 7.7 | 2/3 | 4.60s |
| #38 | GPT-5.4 Nano medium | OpenAI | 1 | 4.0 | 0/3 | 3.65s |
| #39 | Seed-2.0-Mini medium | Bytedance Seed | 1 | 8.2 | 2/3 | 25.9s |
| #41 | MiMo-V2-Flash medium | Xiaomi | 1 | 7.7 | 2/3 | 3.77s |
| #43 | Qwen3.5-35B-A3B medium | Qwen | 1 | 6.4 | 1/3 | 31.6s |
| #45 | GPT-5 Mini medium | OpenAI | 1 | 5.6 | 1/3 | 14.1s |