Benchmark Radar
AI BENCHMARK PROFILE

TextRich

General AIMultimodal Perception

TextRich evaluates detection of AI-generated text-rich images across six categories including posters, charts, receipts, tables, and UI screenshots, with a released dataset on Hugging Face.

Released
2026-06-17
Readiness
Inspectable
Primary field
General AI

Why it matters

Addresses the gap in detecting synthetic images with heavy text content, which is critical for digital trust and content authenticity.

Motivation

Text-rich images often contain privacy-sensitive, transactional, or decision-relevant information.

Primary resources

Benchmark Radar records only publicly supported details and links back to primary sources for verification.