MEGA-Bench Task - Photo Sharing Image Retrieval: leaderboard

The "Photo Sharing Image Retrieval" 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. 44 models tracked.

Top models

#ModelScoreOverall rank
1Gemini 1.5 Pro (002)100#349
2Gemini 2.5 Pro (Preview 03-25)100#123
3InternVL2.5-78B92.9#431
4InternVL3-38B85.7#405
5InternVL2-8B78.6#817
6Qwen 2 VL 72B78.6#460
7GLM-4.1V-9B (Thinking)78.6#472 (GLM-4.1V-9B)
8Gemma 3 27B (IT)71.4#523
9Gemini 2.0 Flash (Preview)71.4#370
10GPT-4o64.3#343
11Claude 3.5 Sonnet (20240620)64.3#299
12Gemma 3 12B (IT)57.1#664
13Claude 3.5 Sonnet (20241022)57.1#291
14Gemini 1.5 Flash (002)57.1#583
15InternVL3-14B50#505

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-photo-sharing-image-retrieval · How It Works · Data refreshed daily, snapshot 2026-10-20.