OpenCompass Language - Instruction Following — leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 23 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.6 Max Preview | 80 |
| 2 | Gemini 3.1 Pro (Preview) | 79.8 |
| 3 | GPT-5.4 (High) | 79.2 |
| 4 | Hy3-preview (High) | 76 |
| 5 | Qwen 3.5 397B A17B | 75.8 |
| 6 | GPT-5.4 Mini (High) | 74.5 |
| 7 | GLM-5.1 | 74 |
| 8 | Kimi K2.6 | 72 |
| 9 | Qwen 3.6 27B | 69 |
| 10 | Gemma 4 31B (IT) | 68.2 |
| 11 | Claude Opus 4.7 (High) | 65 |
| 12 | Gemini 3.1 Flash Lite (Preview) | 64.2 |
| 13 | DeepSeek V4 Pro | 63.8 |
| 14 | Claude Sonnet 4.6 (High) | 62.8 |
| 15 | MiniMax-M2.7 | 61 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-language-instruction-following · How the rankings work · Data refreshed daily, snapshot 2026-07-22.