Skip to Main Content

Digital Humanities: Transkribus

A guide on DH and DH support at the Radboud University Library


Transkribus is a tool that can be used to generate automatic transcriptions of handwritten (historical) documents. To do this, it uses machine learning based recognition models. You can either use pre-existing text recognition models for a certain type of handwriting, or train your own recognition model to recognize the specific type of writing you are working with. Especially if you train your own models, Transkribus can achieve good results with relatively low error rates.

Besides automatic transcription, it can also be used as a platform for manual transcription and collaborative transcription. Moreover, new features are currently being developed, such as recognition of tables and other complex lay-outs, and a publishing platform for transcribed documents.

Transkribus is developed and offered by the READ-COOP, a cooperative company of which the Radboud University is a member.

With a non-paid Transkribus account, you can use 100 free credits monthly, which corresponds to the automatic recognition of 100 pages of handwritten text. Additional credits need to be bought. You may be able to receive free credits from the University Library (see box 'Pilot 2023-2024' below).

Pilot 2023-2024: free credits and advanced functionalities


Transkribus can be very helpful for academic research in varying disciplines, mainly in (but certainly not limited to) the humanities.

Currently, the University Library is running a pilot with Transkribus, to see whether the tool meets the expectations and needs of our researchers and students. For this, we have taken out an institutional subscription to Transkribus.

Free credits

All academic staff and students of the Radboud University can receive free credits for Transkribus during the academic year of 2023-2024:

  • Everyone can receive up to a maximum of 1000 credits, to be used for Radboud University related studies and projects.
  • In return for receiving these credits, we will ask you to fill in an evaluation form about the tool afterwards

Would you like to make use of these free credits? Then send an e-mail to with the amount of credits you need and what you will use it for. Credits can be transferred to any Transkribus-account that uses an mail address.

Advanced functionalities

Certain features and new functionalities of Transkribus are only available with paid subscriptions. These include table recognition, 'super models' for text recognition, more storage, and faster processing speed (See also: Our institutional subscription to Transkribus allows for a limited number of users making use of these paid features. Would you like to gain access to these features, please let us know at We will ask you to fill in an evaluation form about the tool afterwards.

After the pilot

After the summer of 2024, the University Library will evaluate this pilot and decide whether to maintain an institutional subscription on a structural basis.

A short introduction to Transkribus

This is a recording of an online demonstration of Transkribus presented on September 21, 2023, by Digital Humanities Librarian Levi Damsma.

Privacy and personal data

Documents uploaded to Transkribus are hosted on the READ-COOP servers in Innsbruck, Austria. That means that, technically, you will be sharing the information in those documents with the READ-COOP. Do you want to use Transkribus to work with documents of up to 100 years old containing personal data? Then first discuss this with the Privacy Officer of your faculty.

The READ-COOP complies with EU law including the GDPR (known in Dutch as AVG) and you retain full ownership and all rights to the data that you upload.