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。 排序方式: 响应时间（平均） ↓.
455/455
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） |
|---|---|---|---|---|---|---|---|
| #50 | Gemini 4 Argon medium | 1.0 | 0.9 | $0.674 | 2/2 | 2.71s | |
| #124 | Gemini 3.5 Flash medium | 1.0 | 0.8 | $1.316 | 2/2 | 2.70s | |
| #336 | Kimi K2.5 none | Moonshot AI | 0.7 | 0.5 | $0.262 | 1/2 | 2.67s |
| #238 | KAT-Coder-Pro V2.5 high | Kwaipilot | 1.0 | 0.6 | $0.506 | 2/2 | 2.67s |
| #14 | GPT-6 Astra low | OpenAI | 1.0 | 1.0 | $2.160 | 2/2 | 2.65s |
| #145 | Muse Wonder xhigh | Meta | 0.7 | 0.8 | $4.012 | 1/2 | 2.64s |
| #384 | Owl Alpha default | Openrouter | 0.6 | 0.5 | $0.000 | 1/2 | 2.63s |
| #167 | Gemini 2.5 Flash medium | 1.0 | 0.7 | $0.680 | 2/2 | 2.62s | |
| #284 | KAT-Coder-Pro V2.5 default | Kwaipilot | 1.0 | 0.6 | $0.742 | 2/2 | 2.61s |
| #245 | Claude Sonnet 4.6 medium | Anthropic | 1.0 | 0.6 | $2.126 | 2/2 | 2.61s |
| #198 | Gemini 3.1 Flash Lite medium | 1.0 | 0.7 | $0.153 | 2/2 | 2.59s | |
| #295 | Claude Sonnet 5 none | Anthropic | 0.6 | 0.6 | $1.095 | 1/2 | 2.58s |
| #271 | KAT-Coder-Pro V2.5 medium | Kwaipilot | 1.0 | 0.6 | $0.669 | 2/2 | 2.58s |
| #240 | Laguna XS 2.1 medium | Poolside | 1.0 | 0.6 | $0.083 | 2/2 | 2.57s |
| #173 | Qwen3.8 27B medium | Qwen | 1.0 | 0.7 | ~$0.073 | 2/2 | 2.54s |