谜题求解：API 错误
谜题求解
API 错误
看看哪些 AI 模型在 谜题求解 上最容易遇到 API 错误，更快找出薄弱点。 排序方式: 测试正确 ↑.
23/23
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） |
|---|---|---|---|---|---|---|---|
| #412 | Laguna M.1 default | Poolside | 1 | 0.3 | $0.009 | 0/3 | 891ms |
| #417 | Command A+ medium | Cohere | 3 | 0.3 | $0.064 | 0/3 | 92.1s |
| #423 | Command A+ high | Cohere | 3 | 0.3 | $0.079 | 0/3 | 86.2s |
| #426 | Command A+ none | Cohere | 3 | 0.3 | $0.000 | 0/3 | 88.1s |
| #427 | Command A+ low | Cohere | 3 | 0.3 | $0.073 | 0/3 | 95.8s |
| #429 | Nemotron 3 Nano Omni 30b A3b Reasoning medium | NVIDIA | 1 | 0.3 | $0.000 | 0/3 | 1.40s |
| #431 | Nemotron 3 Nano Omni 30b A3b Reasoning default | NVIDIA | 1 | 0.3 | $0.000 | 0/3 | 532ms |
| #445 | LFM2-24B-A2B default | Liquid | 1 | 0.4 | $0.001 | 0/3 | 1.78s |
| #41 | Muse Wonder low | Meta | 1 | 0.6 | $2.637 | 1/3 | 5.08s |
| #44 | Muse Wonder minimal | Meta | 2 | 0.5 | $1.639 | 1/3 | 1.36s |
| #134 | Muse Wonder high | Meta | 2 | 0.5 | $3.426 | 1/3 | 3.78s |
| #142 | Muse Wonder xhigh | Meta | 2 | 0.5 | $4.012 | 1/3 | 3.53s |
| #383 | Hy3 preview low | Tencent | 1 | 0.5 | $0.042 | 1/3 | 7.51s |
| #398 | Qwen3.6 Plus Preview medium | Qwen | 2 | 0.5 | $0.000 | 1/3 | 7.52s |
| #403 | Laguna M.1 medium | Poolside | 1 | 0.5 | $0.033 | 1/3 | 10.2s |