MEGA-Bench - Skill: Language Understanding And Generation: leaderboard
Metric: Average Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 70.2 |
| 2 | GPT-4o | 61.6 |
| 3 | Claude 3.5 Sonnet (20241022) | 60.8 |
| 4 | Claude 3.5 Sonnet (20240620) | 57.5 |
| 5 | Gemini 1.5 Pro (002) | 55.3 |
| 6 | Qwen 2 VL 72B | 53.7 |
| 7 | GPT-4o Mini | 51.7 |
| 8 | InternVL3-78B | 50.8 |
| 9 | Gemini 1.5 Flash (002) | 49.8 |
| 10 | InternVL3-38B | 49.3 |
| 11 | InternVL2.5-78B | 49.3 |
| 12 | Gemma 3 27B (IT) | 47.2 |
| 13 | Gemma 3 12B (IT) | 44.4 |
| 14 | InternVL3-8B | 41.2 |
| 15 | Llama 4 Scout Base | 40.9 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-skill-language-understanding-and-generation · How It Works · Data refreshed daily, snapshot 2026-09-05.