TextFake - Screenshots: leaderboard

Metric: Accuracy (%) on the screen-capture images (80% of real and of generated images) at classifying text-rich images (screenshots, documents and news pages in 28 languages) as real or AI-generated; the balanced set pairs real images with fakes from four text-rendering image generators, every image JPEG-re-encoded; VLM APIs answer REAL or FAKE to one fixed zero-shot prompt; chance 50; higher is better. Source: arxiv.org. Saturation forecast: Around December 2026. 3 models tracked.

Top models

#ModelScore
1Gemini 3 Pro (Preview)75.32
2Claude Sonnet 4.672.1
3GPT-5.464.74

Interactive version: theaggregate.ai/benchmark?slug=textfake-screenshots · How It Works · Data refreshed daily, snapshot 2026-09-29.