反AI技巧：答案错误
反AI技巧
答案错误
看看哪些 AI 模型在 反AI技巧 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 ↑.
249/249
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） |
|---|---|---|---|---|---|---|---|
| #13 | GPT-5.6 Sol low | OpenAI | 1 | 0.8 | $1.244 | 3/4 | 2.60s |
| #22 | GPT-5.6 Sol high | OpenAI | 1 | 0.9 | $1.595 | 3/4 | 3.39s |
| #29 | Seed 2.1 Turbo low | Bytedance Seed | 1 | 0.8 | $1.645 | 3/4 | 34.8s |
| #37 | GPT-5.3-Codex medium | OpenAI | 1 | 0.9 | $1.318 | 3/4 | 4.16s |
| #42 | Seed 2.1 Turbo high | Bytedance Seed | 1 | 0.8 | $1.892 | 3/4 | 47.0s |
| #51 | Gemini 3.6 Flash low | 1 | 0.9 | $0.363 | 3/4 | 1.85s | |
| #60 | GPT-5.4 medium | OpenAI | 1 | 0.8 | $2.213 | 3/4 | 4.11s |
| #61 | Muse Spark Open medium | Meta | 1 | 0.7 | $1.260 | 2/4 | 175.4s |
| #62 | DeepSeek V4.1 Flash high | DeepSeek | 1 | 0.8 | $0.617 | 3/4 | 3.42s |
| #63 | DeepSeek V4.1 Flash medium | DeepSeek | 1 | 0.8 | $0.591 | 3/4 | 3.07s |
| #65 | GPT-5.2 medium | OpenAI | 1 | 0.7 | $1.558 | 2/4 | 7.81s |
| #67 | MiMo-V2.6-Pro medium | Xiaomi | 1 | 0.7 | $0.586 | 2/4 | 2.76s |
| #68 | Seed 2.1 Turbo medium | Bytedance Seed | 1 | 0.8 | $2.042 | 3/4 | 28.2s |
| #70 | DeepSeek V4.1 Flash max | DeepSeek | 1 | 0.8 | $1.204 | 3/4 | 4.35s |
| #74 | Claude Haiku 5.5 low | Anthropic | 1 | 0.9 | $0.061 | 3/4 | 2.16s |