OpenCompass Language - NLP — leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 23 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 (High) | 74.6 |
| 2 | Qwen 3.6 Max Preview | 72.9 |
| 3 | GPT-5.4 Mini (High) | 69.6 |
| 4 | Gemini 3.1 Pro (Preview) | 67.5 |
| 5 | Kimi K2.6 | 67.5 |
| 6 | Claude Sonnet 4.6 (High) | 66.2 |
| 7 | Claude Opus 4.7 (High) | 65.4 |
| 8 | GLM-5.1 | 63.8 |
| 9 | DeepSeek V4 Pro | 63.3 |
| 10 | MiniMax-M2.7 | 63.3 |
| 11 | Qwen 3.6 27B | 63.3 |
| 12 | Hy3-preview (High) | 63.3 |
| 13 | Qwen 3.5 397B A17B | 62.5 |
| 14 | DeepSeek V4 Flash | 62.5 |
| 15 | JT-MINI | 62.1 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-language-nlp · How the rankings work · Data refreshed daily, snapshot 2026-07-22.