指令遵循 排名
看看哪些 AI 模型在 指令遵循 上表现最好，哪些更稳定，以及差距主要出现在哪里。
311/311
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 | 指令遵循 得分 | 分数 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #1 | Gemini 3.6 Flash high | 10.0 | 9.9 | $0.699 | 2/2 | 4.07s | |
| #2 | Gemini 3.8 Flash high | 10.0 | 9.9 | $1.595 | 2/2 | 4.73s | |
| #3 | Gemini 3.7 Flash high | 10.0 | 9.8 | $0.646 | 2/2 | 3.77s | |
| #4 | Gemini 3.6 Flash medium | 10.0 | 9.8 | $0.427 | 2/2 | 3.82s | |
| #5 | GPT-5.6 Sol low | OpenAI | 10.0 | 9.5 | $0.357 | 2/2 | 2.27s |
| #6 | Gemini 3 Flash Preview medium | 10.0 | 9.5 | $0.763 | 2/2 | 4.04s | |
| #7 | Gemini 3.5 Flash high | 10.0 | 9.4 | $2.011 | 2/2 | 3.35s | |
| #8 | Gemini 3.7 Flash medium | 10.0 | 9.4 | $0.309 | 2/2 | 2.81s | |
| #9 | GPT-5.6 Sol medium | OpenAI | 10.0 | 9.4 | $0.508 | 2/2 | 2.50s |
| #10 | GPT-5.6 Sol high | OpenAI | 10.0 | 9.4 | $0.446 | 2/2 | 2.73s |
| #13 | Gemini 3.1 Pro Preview medium | 10.0 | 9.2 | $1.352 | 2/2 | 9.56s | |
| #14 | Gemini 3.7 Flash low | 10.0 | 9.2 | $0.207 | 2/2 | 2.47s | |
| #15 | Claude Opus 5 high | Anthropic | 10.0 | 9.2 | $3.070 | 2/2 | 4.78s |
| #16 | Claude Opus 5 medium | Anthropic | 10.0 | 9.2 | $1.586 | 2/2 | 4.16s |
| #17 | Gemini 3.8 Flash medium | 10.0 | 9.1 | $0.653 | 2/2 | 4.81s |