反AI技巧：API 错误
反AI技巧
API 错误
看看哪些 AI 模型在 反AI技巧 上最容易遇到 API 错误，更快找出薄弱点。 排序方式: 测试正确 ↓.
31/31
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 | API 错误 次数 | 分类得分 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #32 | Muse Wonder medium | Meta | 1 | 0.8 | $3.397 | 3/4 | 4.84s |
| #131 | Muse Wonder high | Meta | 1 | 0.8 | $3.426 | 3/4 | 5.93s |
| #139 | Muse Wonder xhigh | Meta | 1 | 0.8 | $4.012 | 3/4 | 6.04s |
| #184 | Ember-1 low | Fireworks | 1 | 0.8 | $1.488 | 3/4 | 22.2s |
| #191 | Ember-1 high | Fireworks | 1 | 0.9 | $2.405 | 3/4 | 24.0s |
| #203 | Grok 4.7 xhigh | X AI | 1 | 0.9 | $3.398 | 3/4 | 35.4s |
| #207 | Grok 4.7 high | X AI | 1 | 0.9 | $3.037 | 3/4 | 28.0s |
| #209 | Ember-1 max | Fireworks | 1 | 0.9 | $3.264 | 3/4 | 23.7s |
| #255 | DeepSeek V3.2 medium | DeepSeek | 1 | 0.8 | $0.152 | 3/4 | 24.2s |
| #374 | Hy3 preview low | Tencent | 1 | 0.8 | $0.042 | 3/4 | 9.32s |
| #389 | Qwen3.6 Plus Preview medium | Qwen | 1 | 0.8 | $0.000 | 3/4 | 11.7s |
| #141 | GPT-6 Luna low | OpenAI | 1 | 0.7 | $0.033 | 2/4 | 11.0s |
| #185 | Step 5 Preview high | Stepfun | 1 | 0.7 | $4.513 | 2/4 | 14.7s |
| #271 | Step 5 Preview medium | Stepfun | 1 | 0.6 | $4.136 | 2/4 | 12.6s |
| #302 | Step 5 Preview low | Stepfun | 1 | 0.7 | $4.110 | 2/4 | 9.50s |