领域专项：答案错误
领域专项
答案错误
看看哪些 AI 模型在 领域专项 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 ↑.
299/299
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） |
|---|---|---|---|---|---|---|---|
| #1 | Gemini 3.6 Flash high | 1 | 8.2 | $0.699 | 2/3 | 33.3s | |
| #6 | Gemini 3 Flash Preview medium | 1 | 7.7 | $0.763 | 2/3 | 20.1s | |
| #8 | Gemini 3.7 Flash medium | 1 | 7.7 | $0.309 | 2/3 | 6.49s | |
| #13 | Gemini 3.1 Pro Preview medium | 1 | 7.7 | $1.352 | 2/3 | 20.3s | |
| #14 | Gemini 3.7 Flash low | 1 | 8.2 | $0.207 | 2/3 | 3.09s | |
| #17 | Gemini 3.8 Flash medium | 1 | 8.2 | $0.653 | 2/3 | 31.8s | |
| #18 | Seed 2.1 Turbo low | Bytedance Seed | 1 | 2.9 | $1.546 | 0/3 | 424.5s |
| #20 | Claude Fable 5.1 high | Anthropic | 1 | 7.7 | $3.364 | 2/3 | 24.3s |
| #33 | Gemini 3.6 Flash low | 1 | 7.6 | $0.251 | 2/3 | 4.65s | |
| #39 | Claude Fable 5 medium | Anthropic | 1 | 5.3 | $3.004 | 1/3 | 37.3s |
| #47 | GPT-5.2 medium | OpenAI | 1 | 5.9 | $1.182 | 1/3 | 97.3s |
| #48 | Qwen3.8 27B high | Qwen | 1 | 5.3 | ~$0.287 | 1/3 | 526.7s |
| #52 | Claude Fable 5.1 medium | Anthropic | 1 | 7.7 | $2.909 | 2/3 | 19.3s |
| #59 | Inkling high | Thinkingmachines | 1 | 7.7 | $1.046 | 2/3 | 174.7s |
| #60 | Muse Spark 1.1 high | Meta | 1 | 2.9 | $2.253 | 0/3 | 94.7s |