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: 分数 ↑.
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） |
|---|---|---|---|---|---|---|---|
| #30 | GPT-5.3-Codex medium | OpenAI | 4 | 8.9 | $0.988 | 16/22 | 16.9s |
| #29 | Grok 4.5 high | X AI | 2 | 9.0 | $2.252 | 17/22 | 72.3s |
| #28 | Qwen3.8 Max medium | Qwen | 4 | 9.0 | $0.771 | 17/22 | 23.3s |
| #27 | GPT-5.5 medium | OpenAI | 4 | 9.0 | $4.163 | 18/22 | 39.0s |
| #26 | Gemini 3.5 Flash medium | 3 | 9.0 | $0.665 | 18/22 | 8.48s | |
| #25 | Claude Fable 5.1 high | Anthropic | 3 | 9.0 | $3.364 | 19/22 | 13.6s |
| #24 | Qwen3.7 Max medium | Qwen | 4 | 9.1 | $1.157 | 17/22 | 44.3s |
| #23 | Seed 2.1 Turbo low | Bytedance Seed | 3 | 9.1 | $1.546 | 17/22 | 163.9s |
| #22 | Gemini 3.8 Flash medium | 2 | 9.1 | $0.653 | 19/22 | 14.5s | |
| #21 | Claude Opus 5 medium | Anthropic | 3 | 9.2 | $1.586 | 18/22 | 10.3s |
| #20 | Claude Opus 5 high | Anthropic | 2 | 9.2 | $3.070 | 18/22 | 22.5s |
| #19 | Gemini 3.7 Flash low | 2 | 9.2 | $0.207 | 20/22 | 3.85s | |
| #18 | Gemini 3.1 Pro Preview medium | 2 | 9.2 | $1.352 | 20/22 | 20.6s | |
| #17 | Grok 4.6 high | X AI | 2 | 9.3 | $1.908 | 19/22 | 84.1s |
| #16 | GPT-5.5 low | OpenAI | 3 | 9.3 | $1.257 | 19/22 | 10.1s |