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。
443/443
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） |
|---|---|---|---|---|---|---|---|
| #261 | Claude Opus 4.6 medium | Anthropic | 4 | 0.6 | $2.961 | 13/22 | 32.2s |
| #269 | GLM 5 Turbo medium | Z.ai | 4 | 0.6 | $0.323 | 14/21 | 23.0s |
| #272 | Claude Fable 5.1 low | Anthropic | 4 | 0.6 | $3.395 | 11/23 | 11.4s |
| #273 | Gemini 3.5 Flash none | 4 | 0.6 | $1.752 | 14/23 | 14.1s | |
| #375 | Gemini 3.1 Flash Lite high | 4 | 0.5 | $2.044 | 10/18 | 62.0s | |
| #383 | Hy3 preview low | Tencent | 4 | 0.5 | $0.042 | 10/21 | 24.6s |
| #400 | Grok 4.20 Multi Agent Beta medium | X AI | 4 | 0.4 | $5.599 | 8/18 | 9.69s |
| #401 | Hunter Alpha medium | OpenRouter | 4 | 0.4 | $0.000 | 8/18 | 10.3s |
| #402 | Grok 4.1 Fast medium | X AI | 4 | 0.4 | $0.069 | 9/19 | 23.8s |
| #403 | Laguna M.1 medium | Poolside | 4 | 0.4 | $0.033 | 9/19 | 14.7s |
| #446 | Solar Decide Flash default | Upstage | 4 | 0.2 | $0.003 | 3/7 | 1.33s |
| #449 | Tev1 4B Experimental default | Togethercomputer | 4 | 1.0 | $0.002 | 3/7 | 498ms |
| #21 | GPT-6 Astra medium | OpenAI | 3 | 0.9 | $2.707 | 20/23 | 9.90s |
| #23 | GPT-5.5 low | OpenAI | 3 | 0.9 | $1.822 | 20/23 | 11.9s |
| #28 | Claude Opus 5 medium | Anthropic | 3 | 0.9 | $2.870 | 19/23 | 13.2s |