MEGA-Bench Task - Physics Exams V: leaderboard

The "Physics Exams V" subtask of MEGA-Bench Task. A single task from MEGA-Bench, a large-scale multimodal evaluation suite that scores vision-language models across 500+ real-world tasks spanning perception, reasoning, OCR, charts, documents, and multi-image understanding.

Metric: Task Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 17 models tracked.

Top models

#ModelScoreOverall rank
1Gemini 2.5 Pro (Preview 03-25)78.6#123
2InternVL3-38B57.1#405
3Gemini 2.0 Flash (Preview)42.9#370
4InternVL3-8B35.7#620
5InternVL3-78B35.7#352
6GLM-4.1V-9B (Thinking)35.7#472 (GLM-4.1V-9B)
7Gemma 3 4B (IT)28.6#992
8InternVL3-14B28.6#505
9Gemma 3 27B (IT)21.4#523
10Gemma 3 12B (IT)7.1#664
11Llama 4 Scout Base7.1#707

No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-physics-exams-v · How It Works · Data refreshed daily, snapshot 2026-10-20.