MEGA-Bench Task - NExT-QA OE: leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 47.4 |
| 2 | GPT-4o | 44.2 |
| 3 | InternVL2.5-78B | 40 |
| 4 | InternVL3-78B | 37.4 |
| 5 | Claude 3.5 Sonnet (20240620) | 36.8 |
| 6 | Claude 3.5 Sonnet (20241022) | 36.3 |
| 7 | Aquila-VL-2B | 35.8 |
| 8 | Llama 4 Scout Base | 35.8 |
| 9 | Gemma 3 27B (IT) | 35.3 |
| 10 | Gemini 1.5 Pro (002) | 35.3 |
| 11 | Pixtral-12B | 35.3 |
| 12 | InternVL3-8B | 34.7 |
| 13 | Gemma 3 12B (IT) | 34.2 |
| 14 | GPT-4o Mini | 33.7 |
| 15 | InternVL3-38B | 32.6 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-next-qa-oe · How It Works · Data refreshed daily, snapshot 2026-09-05.