Extract English printed text and optionally add a searchable layer to scanned PDF pages.
PDF rendering, OCR, and searchable-PDF creation happen in this browser. The worker, WASM core, and English model are self-hosted; no CDN or OCR service receives your document.
English printed text
The first run loads roughly 7 MB of local OCR assets and caches the English model in this browser. Handwriting and decorative fonts are not reliable.
Review important text
OCR is probabilistic and can misread names, numbers, and tables. Confidence is a clue, not proof of accuracy.
Extract PDF Fonts
Recover embedded TrueType font programs from selected PDF pages locally with licensing and subset warnings.
Extract PDF Images
Recover original embedded raster images from selected PDF pages locally without rendering page screenshots.
Markdown to PDF
Typeset GitHub-flavoured Markdown as a PDF — tables, task lists, and code blocks included — with a choice of themes.
PDF Accessibility Structure Checker
Preflight PDF tags, language, titles, figures, fields, and links locally without claiming certification.