反AI技巧 排名
看看哪些 AI 模型在 反AI技巧 上表现最好，哪些更稳定，以及差距主要出现在哪里。 排序方式: 测试正确 ↑.
461/461
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 | 反AI技巧 得分 | 分数 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #277 | KAT-Coder-Pro V2.5 medium | Kwaipilot | 0.8 | 0.6 | $0.669 | 3/4 | 3.90s |
| #283 | Beam 1.0 501B A23B high | Reflection | 0.9 | 0.6 | $0.674 | 3/4 | 1.36s |
| #285 | Apodex 1.1 Mini low | Apodex | 0.9 | 0.6 | $0.000 | 3/4 | 9.10s |
| #289 | Granite 4.2 8B medium | IBM Granite | 0.8 | 0.6 | $0.029 | 3/4 | 10.3s |
| #290 | KAT-Coder-Pro V2.5 default | Kwaipilot | 0.9 | 0.6 | $0.742 | 3/4 | 4.58s |
| #293 | GLM 5.3 FlashX low | Z.ai | 0.8 | 0.6 | $0.157 | 3/4 | 1.56s |
| #297 | Gemini 3.1 Flash Lite minimal | 0.8 | 0.6 | $0.080 | 3/4 | 1.10s | |
| #315 | Dots 3 Note Preview medium | Dots Studio | 0.9 | 0.6 | $0.000 | 3/4 | 7.76s |
| #324 | Nemotron 3 Super medium | NVIDIA | 0.8 | 0.6 | $0.081 | 3/4 | 7.85s |
| #329 | Gemma 4 26B A4B none | 0.8 | 0.5 | $0.031 | 3/4 | 1.28s | |
| #332 | Claude Opus 4.7 none | Anthropic | 0.8 | 0.5 | $0.505 | 3/4 | 2.12s |
| #334 | North Mini Code medium | Cohere | 0.8 | 0.5 | $0.000 | 3/4 | 64.8s |
| #341 | Granite 4.2 8B low | IBM Granite | 0.8 | 0.5 | $0.036 | 3/4 | 11.7s |
| #346 | Inkling Small low | Thinkingmachines | 0.9 | 0.5 | $0.055 | 3/4 | 1.81s |
| #351 | MiMo-V2-Flash medium | Xiaomi | 0.8 | 0.5 | $0.043 | 3/4 | 15.8s |