AEGIS (Academic Image Forensics) - Textual Artifact Recognition: leaderboard

AEGIS (2026) tests whether multimodal models can audit AI-generated or AI-edited images in academic papers, from figure panels across seven scientific image categories and forgeries made with 25 image generators. Macro-F1 for deciding whether the text inside an image shows AI-synthesis traces, in percent; higher is better.

Metric: Macro-F1 (%) over two classes, telling whether the text regions of an image carry AI-synthesis traces, on AEGIS academic images (seven categories, 39 subtypes; real panels and forgeries from 25 generative models under four forgery strategies), zero-shot minimal prompts, lossless PNG input; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 26 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 3 Pro (Preview)84.67#119
2GPT-4.182.74#505
3Gemini 2.5 Flash80.97#463
4O4 MiniHigh80.02#368
5Qwen 3 VL Plus78.98#512
6Doubao-Seed-1.6Thinking77.57
7Seed-1.676.9#469
8GPT-5.176.38#240
9Doubao-1.5-thinking-vision-pro75.44#344
10Qwen-VL-Max (AEGIS checkpoint unspecified)73.74
11Doubao-1.5 Vision Pro72.32#1026
12Qwen 2.5 VL 72B Instruct70.76#841
13Llama 4 Maverick68.32#1110
14Ministral 3 14B62.25#1870
15Doubao-Seed-1.6-Flash60.39#1250

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=aegis-academic-image-forensics-textual-artifact-recognition · How It Works · Data refreshed daily, snapshot 2026-11-21.