答案错误 失败
看看哪些 AI 模型最常遇到 答案错误，让你在选择前先发现稳定性风险。 排序方式: 测试正确 ↑.
215/215
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 | 分数 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #57 | GPT-5.4 Nano medium | OpenAI | 8 | 7.5 | $0.138 | 12/22 | 13.2s |
| #60 | GPT-5.4 Mini medium | OpenAI | 6 | 7.5 | $0.756 | 12/22 | 25.9s |
| #64 | LongCat 2.0 medium | Meituan | 7 | 7.4 | $0.478 | 12/22 | 136.6s |
| #67 | Claude Sonnet 4.6 none | Anthropic | 5 | 7.3 | $0.661 | 12/22 | 8.12s |
| #71 | Step 3.7 Flash low | Stepfun | 8 | 7.3 | $0.454 | 12/22 | 20.7s |
| #72 | Kimi K2.6 medium | Moonshot AI | 3 | 7.2 | $1.036 | 12/22 | 110.0s |
| #79 | Grok 4.20 medium | X AI | 6 | 7.1 | $0.777 | 12/22 | 29.5s |
| #88 | MiMo-V2.5-Pro medium | Xiaomi | 3 | 6.9 | $0.187 | 12/22 | 33.9s |
| #89 | Qwen3.6 Flash medium | Qwen | 8 | 6.9 | $0.738 | 12/22 | 44.7s |
| #95 | Gemini 3.5 Flash-Lite low | 9 | 6.7 | $0.145 | 12/22 | 2.25s | |
| #101 | GLM 5.2 none | Z.ai | 8 | 6.6 | $0.128 | 12/22 | 9.34s |
| #103 | Qwen3.6 Max Preview none | Qwen | 10 | 6.6 | $0.231 | 12/22 | 7.82s |
| #104 | Gemini 3.5 Flash-Lite medium | 9 | 6.5 | $0.369 | 12/22 | 6.01s | |
| #107 | MiMo-V2.5 medium | Xiaomi | 5 | 6.5 | $0.082 | 12/22 | 32.2s |
| #111 | Gemini 3.1 Flash Lite low | 9 | 6.5 | $0.621 | 12/22 | 16.3s |