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。
313/313
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） |
|---|---|---|---|---|---|---|---|
| #203 | Gemini 3 PRO Preview medium | 3 | 6.0 | $0.385 | 14/21 | 9.05s | |
| #210 | Hy3 preview high | Tencent | 3 | 5.9 | $0.135 | 11/21 | 56.6s |
| #233 | Hy4 preview low | Tencent | 3 | 5.6 | $1.745 | 3/22 | 214.3s |
| #6 | Gemini 3.7 Flash high | 2 | 9.8 | $0.646 | 20/22 | 12.2s | |
| #7 | Gemini 3.6 Flash medium | 2 | 9.8 | $0.427 | 20/22 | 11.9s | |
| #9 | GPT-6 Astra low | OpenAI | 2 | 9.5 | $1.289 | 20/22 | 5.09s |
| #10 | Gemini 3 Flash Preview medium | 2 | 9.5 | $0.763 | 20/22 | 19.9s | |
| #11 | Gemini 3.5 Flash high | 2 | 9.4 | $2.011 | 19/22 | 15.5s | |
| #12 | Gemini 3.7 Flash medium | 2 | 9.4 | $0.309 | 20/22 | 5.87s | |
| #17 | Grok 4.6 high | X AI | 2 | 9.3 | $1.908 | 19/22 | 84.1s |
| #18 | Gemini 3.1 Pro Preview medium | 2 | 9.2 | $1.352 | 20/22 | 20.6s | |
| #19 | Gemini 3.7 Flash low | 2 | 9.2 | $0.207 | 20/22 | 3.85s | |
| #20 | Claude Opus 5 high | Anthropic | 2 | 9.2 | $3.070 | 18/22 | 22.5s |
| #22 | Gemini 3.8 Flash medium | 2 | 9.1 | $0.653 | 19/22 | 14.5s | |
| #29 | Grok 4.5 high | X AI | 2 | 9.0 | $2.252 | 17/22 | 72.3s |