OpenCompass Research - IFEval — leaderboard
Metric: Score (%). Source: rank.opencompass.org.cn. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 (High) | 94.6 |
| 2 | Kimi K2.5 | 93.9 |
| 3 | GLM-5 | 93.2 |
| 4 | Step 3.5 Flash | 93.2 |
| 5 | O4 Mini (2025-04-16) (High) | 93 |
| 6 | Gemini 3 Pro (Preview) | 92.8 |
| 7 | GPT-5 Mini (High) | 92.8 |
| 8 | Kimi K2 (Thinking) | 92.4 |
| 9 | O3 (2025-04-16) | 92.2 |
| 10 | DeepSeek V3.2 Speciale | 91.7 |
| 11 | Qwen 3.5 397B A17B | 91.5 |
| 12 | MiniMax-M2.5 | 91.1 |
| 13 | GLM-4.7 | 90.2 |
| 14 | GPT-5 Nano (High) | 90.2 |
| 15 | GPT-OSS-120B (High) | 90.2 |
Interactive version: theaggregate.ai/benchmark?slug=opencompass-research-ifeval · How the rankings work · Data refreshed daily, snapshot 2026-07-22.