OpenCompass LLM - Instruction Following - English (CompassBench 2504): leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 28 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemma 3 27B (IT) | 68.9 |
| 2 | Claude 3.7 Sonnet (Thinking) | 67.8 |
| 3 | Doubao-1.5-thinking-pro-250415 | 67.1 |
| 4 | ERNIE-X1-Turbo-32K | 66.8 |
| 5 | DeepSeek R1 | 66.5 |
| 6 | Gemini 2.5 Pro (Preview 05-06) | 65.7 |
| 7 | Qwen 3 235B A22B (Thinking) | 59.4 |
| 8 | Qwen 3 32B (Thinking) | 57.9 |
| 9 | GPT-4.1 Mini | 56.4 |
| 10 | DeepSeek V3 (0324) | 54.8 |
| 11 | O4 Mini (High) | 54.8 |
| 12 | Qwen 3 235B A22B | 54.4 |
| 13 | Grok 3 | 52.3 |
| 14 | MiniMax-Text-01 | 52.3 |
| 15 | Step 2 16K | 51.6 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-llm-instruction-following-english-compassbench-2504 · How It Works · Data refreshed daily, snapshot 2026-09-19.