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） |
|---|---|---|---|---|---|---|---|
| #32 | Muse Wonder medium | Meta | 0.7 | 0.9 | $3.397 | 1/2 | 3.88s |
| #160 | Inkling Small high | Thinkingmachines | 1.0 | 0.7 | $0.398 | 2/2 | 3.94s |
| #278 | Trinity Large Thinking medium | Arcee AI | 0.4 | 0.6 | $0.862 | 0/2 | 3.99s |
| #61 | Qwen3.8 MAX low | Qwen | 1.0 | 0.9 | $1.089 | 2/2 | 4.01s |
| #442 | Mammoth 1 default | Baseten | 0.2 | 0.2 | $0.000 | 0/1 | 4.04s |
| #15 | Gemini 3 Flash Preview medium | 1.0 | 1.0 | $0.876 | 2/2 | 4.04s | |
| #1 | Gemini 3.6 Flash high | 1.0 | 1.0 | $0.900 | 2/2 | 4.07s | |
| #267 | Trinity Large Thinking low | Arcee AI | 0.4 | 0.6 | $0.918 | 0/2 | 4.07s |
| #317 | Nemotron 3.5 Super VL none | NVIDIA | 1.0 | 0.6 | $0.040 | 2/2 | 4.08s |
| #161 | DeepSeek V4 Pro 0423 none | DeepSeek | 0.6 | 0.7 | $0.083 | 1/2 | 4.12s |
| #304 | Trinity Large Thinking high | Arcee AI | 0.6 | 0.6 | $0.794 | 1/2 | 4.12s |
| #51 | Claude Opus 5 low | Anthropic | 1.0 | 0.9 | $2.350 | 2/2 | 4.13s |
| #28 | Claude Opus 5 medium | Anthropic | 1.0 | 0.9 | $2.870 | 2/2 | 4.16s |
| #404 | Hunter Alpha medium | OpenRouter | 1.0 | 0.4 | $0.000 | 2/2 | 4.18s |
| #282 | Step 5 Preview medium | Stepfun | 1.0 | 0.6 | $4.136 | 2/2 | 4.23s |