MEGA-Bench Task - Perception Test Video Character Order: leaderboard

Metric: Task Score (%). Source: huggingface.co. 44 models tracked.

Top models

#ModelScore
1GPT-4o93.8
2Gemma 3 12B (IT)93.8
3Gemini 1.5 Pro (002)93.8
4Gemma 3 27B (IT)87.5
5Gemini 1.5 Flash (002)87.5
6Gemini 2.5 Pro (Preview 03-25)87.5
7InternVL3-78B81.2
8InternVL3-38B81.2
9GPT-4o Mini81.2
10InternVL2.5-78B81.2
11Claude 3.5 Sonnet (20241022)75
12Claude 3.5 Sonnet (20240620)68.8
13InternVL3-8B62.5
14Gemma 3 4B (IT)43.8
15InternVL2-8B43.8

Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-perception-test-video-character-order · How It Works · Data refreshed daily, snapshot 2026-09-05.