Scanned vs. Text Detector

Classifies every page as text-native, likely-scanned, or blank, entirely in this browser tab — reported per page, not just one yes/no for the whole document.