常识问答：答案错误
常识问答
答案错误
看看哪些 AI 模型在 常识问答 上最容易遇到 答案错误，更快找出薄弱点。
259/259
筛选模型
没有模型匹配当前搜索和筛选条件。
| 排名 | 模型 | 公司 | 答案错误 次数 | 分类得分 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #35 | Seed 2.1 Turbo high | Bytedance Seed | 1 | 3.0 | $1.797 | 0/1 | 78.6s |
| #37 | Qwen3.8 Max (0902) medium | Qwen | 1 | 3.0 | $0.677 | 0/1 | 17.0s |
| #38 | Muse Spark 1.3 medium | Meta | 1 | 3.0 | $1.469 | 0/1 | 85.6s |
| #39 | Gemini 3.5 Flash Lite high | 1 | 3.0 | $0.540 | 0/1 | 9.68s | |
| #41 | Claude Opus 5 low | Anthropic | 1 | 3.0 | $1.121 | 0/1 | 6.48s |
| #44 | Claude Opus 4.7 medium | Anthropic | 1 | 3.0 | $1.476 | 0/1 | 2.25s |
| #45 | Muse Spark 1.2 high | Meta | 1 | 3.0 | $1.771 | 0/1 | 15.0s |
| #46 | Muse Spark 1.2 medium | Meta | 1 | 3.0 | $1.339 | 0/1 | 12.8s |
| #47 | Muse Spark 1.1 medium | Meta | 1 | 3.0 | $1.420 | 0/1 | 18.5s |
| #49 | Qwen3.6 Max Preview medium | Qwen | 1 | 3.0 | $1.132 | 0/1 | 60.6s |
| #50 | Qwen3.8 Max low | Qwen | 1 | 3.0 | $0.702 | 0/1 | 12.4s |
| #51 | GPT-5.4 medium | OpenAI | 1 | 3.0 | $1.560 | 0/1 | 14.0s |
| #52 | DeepSeek V4.1 Flash low | DeepSeek | 1 | 3.0 | $0.448 | 0/1 | 46.9s |
| #53 | DeepSeek V4.1 Flash high | DeepSeek | 1 | 3.0 | $0.574 | 0/1 | 38.0s |
| #54 | DeepSeek V4.1 Flash medium | DeepSeek | 1 | 3.0 | $0.544 | 0/1 | 20.3s |