wondarchive

wondarchive › Choosing a tool

Which tool should you use to transcribe archives?

For archive photos taken in the reading room that you need to read, translate and search, wondarchive does it all in a single online service, handwriting included, with nothing to install. Other tools suit other situations: a line-by-line edition with a trained model, clean scans of recent printed material, or a pipeline you build yourself.

Which tool for which collection?

Your situationThe right toolWhy
Thousands of handheld photos taken in the reading room, to read, search and citewondarchivestraightening, transcription, translation, filing by shelfmark and search in a single service, the collection readable the day after upload
A collection in a language you read poorlywondarchiveeach page transcribed in its own language, then translated into yours, with the context of neighbouring pages
Handwritten letters, registers or deedswondarchiveextra-strength reread by our most precise model, with no prior training; 100 free pages to judge
A line-by-line scholarly edition of a single hand, with an engineereScriptorium (open source) or Transkribusfine segmentation and models trained on one hand, ALTO and PAGE export
Early printed books and blackletter, with an open-source pipeline to installOCR4allopen-source software from the University of Würzburg
Clean scans of recent printed material, as PDFsABBYY FineReaderdesktop OCR on clean scans
Organising and describing your photos without reading themTropyfree, on your own computer; no transcription
Millions of pages, a pipeline to buildArkindex or a cloud OCRmass processing, on quote or pay per call, with developers
Translating a text already transcribedDeepLtranslation only; does not read an image
No budgetthe free options100 free pages, transcription on your computer, open-source software

What should you ask yourself before choosing?

  1. What kind of script: typewritten, printed, blackletter, handwritten?
  2. How many pages, and in what form: photos from an archive trip, scans, PDFs?
  3. Does it need translating?
  4. Do you need to search the whole collection and cite a page by its shelfmark and folio?
  5. Where do the images go, and are they used to train a model?
  6. Do you have a server and an engineer to install software?
  7. Do you work alone or in a team?

Why wondarchive for archive photos?

wondarchive is an online service that transcribes, translates and indexes archive photos taken in the reading room, for historians, PhD students, archivists and genealogists. You upload the folder just as the camera wrote it: each page is turned the right way up, read by AI from the full-resolution original, translated if needed, filed by repository, shelfmark and folio, and the whole collection becomes searchable by word, name and date. The images are stored in France, with OVHcloud, read in the EU, and train no model.

When is another tool the better choice?

For a line-by-line edition of a single hand, with ALTO or PAGE export, eScriptorium or Transkribus are built for exactly that. For PDF scans of recent printed material where you only want the text, desktop OCR is enough. To simply organise and describe photos, Tropy is free. The comparison details each tool: who makes it, where, and where your images go.

The account takes a minute to open, a hundred pages free for a year.

What is the best software for transcribing handwritten archives?

For photos of manuscripts to read and search without training, wondarchive: its extra-strength reread reads cursive handwriting in European languages, in the Latin, Cyrillic and Greek alphabets. For a line-by-line edition of a single hand with a trained model, eScriptorium or Transkribus.

Is there an archive transcription tool hosted in France?

Yes: wondarchive is hosted by OVHcloud, in France, and uploaded documents train no model. eScriptorium, developed at the EPHE, installs on your own server.

Do you need to train a model to read handwriting?

Not necessarily any more: recent AI models read the whole page without training on your handwriting. Training is still worthwhile for a line-by-line edition of a single hand.

Can I try before choosing?

Yes: the free account transcribes and translates 100 pages, kept for a year, with no credit card, and the demo opens six real collections without an account.

See it on your own pages

A hundred pages transcribed and translated, for a year, no card required.