ReTyped

Document reconstruction or translation

Rebuild or translate scanned archives.

ReTyped is currently available by request. Explore the public examples to inspect reconstructed PDFs, Markdown, geometry, and editable translations; then request access if you want to test your own pages.

Request accessFor teams

The demos remain public. New document processing is granted manually after a request.

Output examples

ReTyped

Reconstructed PDF

Rebuilt typography, columns, and photos in place.

Searchable PDF

Your scan with an accurate word-level text layer.

Markdown

Clean prose in reading order with semantic zones.

Geometry JSON

Word-level bounding boxes, reading-order zones.

Translate

Translated PDF

The page in the target language, with layout preserved and text still editable.

Translated Markdown

Same structure, in the target language, ready to review or correct.

Spin Magazine #19 (November 1986)

Built for collections, useful from the first page

The public service lets you evaluate the technology on real examples, from one scanned book to a larger archive. The same pipeline can be adapted for bigger collections: private queues, API access, custom outputs, and retention requirements.

  • Reconstructed PDFs with rebuilt columns, justified typography, and images in place: a usable digital document, not just a scan with hidden text.
  • Clean Markdown in reading order, with word-level geometry linked by zone ID for indexing, analysis, or downstream systems.
  • High-quality layout-preserving translation, review, and editing for scanned documents or multilingual collections, with room to discuss API access, batch processing, or dedicated infrastructure.