反AI技巧：额外格式
反AI技巧
额外格式
看看哪些 AI 模型在 反AI技巧 上最容易遇到 额外格式，更快找出薄弱点。
21/21
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 | 额外格式 次数 | 分类得分 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #82 | Claude Opus 4.6 medium | Anthropic | 2 | 6.4 | $2.961 | 2/4 | 7.45s |
| #108 | Claude Sonnet 4.6 none | Anthropic | 2 | 4.8 | $0.661 | 1/4 | 2.94s |
| #115 | Claude Opus 4.8 none | Anthropic | 2 | 6.5 | $1.166 | 2/4 | 3.40s |
| #136 | Claude Fable 5.1 low | Anthropic | 2 | 6.5 | $2.565 | 2/4 | 4.54s |
| #181 | Claude Sonnet 5 none | Anthropic | 2 | 5.3 | $0.548 | 1/4 | 3.60s |
| #258 | North Mini Code none | Cohere | 2 | 3.0 | $0.000 | 0/4 | 22.5s |
| #262 | DeepSeek V3.2 none | DeepSeek | 2 | 3.2 | $0.054 | 0/4 | 9.35s |
| #291 | Granite 4.2 8B none | IBM Granite | 2 | 4.9 | $0.026 | 1/4 | 17.8s |
| #37 | Muse Spark 1.2 medium | Meta | 1 | 7.3 | $1.339 | 2/4 | 5.38s |
| #52 | Claude Fable 5.1 medium | Anthropic | 1 | 8.2 | $2.909 | 3/4 | 5.18s |
| #75 | Claude Sonnet 4.6 medium | Anthropic | 1 | 6.5 | $2.126 | 2/4 | 2.98s |
| #95 | Qwen3.5-27B medium | Qwen | 1 | 8.7 | $0.982 | 3/4 | 19.8s |
| #98 | Grok Build 0.1 medium | X AI | 1 | 8.3 | $1.130 | 3/4 | 7.43s |
| #145 | DeepSeek V4 Pro none | DeepSeek | 1 | 3.2 | $0.226 | 0/4 | 4.02s |
| #177 | MiMo-V2-Flash medium | Xiaomi | 1 | 8.1 | $0.043 | 3/4 | 15.8s |