MEGA-Bench Task - Screenshot LightEval Math — leaderboard
Metric: Task Score (%). Source: huggingface.co. 17 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL3-38B | 73.3 |
| 2 | InternVL3-14B | 73.3 |
| 3 | Gemini 2.5 Pro | 66.7 |
| 4 | Gemini 2.0 Flash (Preview) | 66.7 |
| 5 | InternVL3-78B | 60 |
| 6 | Llama 4 Scout Base | 60 |
| 7 | Gemma 3 27B (IT) | 53.3 |
| 8 | Ovis2-8B | 46.7 |
| 9 | InternVL3-8B | 40 |
| 10 | Gemma 3 12B (IT) | 26.7 |
| 11 | Gemma 3 4B (IT) | 20 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-screenshot-lighteval-math · How the rankings work · Data refreshed daily, snapshot 2026-07-22.