MEGA-Bench Task - Screenshot TheoremQA — leaderboard

Metric: Task Score (%). Source: huggingface.co. 17 models tracked.

Top models

#ModelScore
1Gemini 2.0 Flash (Preview)85.7
2Gemini 2.5 Pro78.6
3Gemma 3 27B (IT)64.3
4InternVL3-78B57.1
5InternVL3-38B50
6InternVL3-14B50
7Gemma 3 12B (IT)42.9
8Llama 4 Scout Base42.9
9InternVL3-8B35.7
10Ovis2-8B21.4
11Gemma 3 4B (IT)14.3

Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-screenshot-theoremqa · How the rankings work · Data refreshed daily, snapshot 2026-07-22.