Transkribus

The NorHand Model: a new public AI model from the National Library of Norway Trained on 400+ hands, the NorHand model helps the National Library of Norway digitise its collection via Transkribus, making cultural heritage accessible.

Transcribing an entire library?

The National Library of Norway has already processed 25% of their scanned documents using Transkribus and the NorHand model.

This massive effort will make centuries of cultural heritage searchable and widely accessible.

👉 blog.transkribus.org/en/the-norha...

📢 Save the Date: TUC 2026 @ University of Passau

From 21–23 September 2026, we’re bringing the Transkribus User Conference to the beautiful University of Passau!

Learn more and stay updated here: tuc.transkribus.org

YouTube video by Transkribus Transkribus Table Models Webinar (English)

Missed our live session on Table Models? You can now catch the full recording on YouTube!

👉Watch the recording here: www.youtube.com/watch?v=NYkG...

Working with difficult layouts? Today's Field Models webinar will show you how to automate the tagging and recognition of marginalia, headers or signatures.

Sign-up is still open! 👉https://www.transkribus.org/event/field-models-january-2026

We were thrilled to see the relaunch of the digital archive of the Arbeiter-Zeitung (Workers' Newspaper) by Medienhaus Wien last autumn!

Transkribus helped to turn 215,000 pages into a fully searchable digital resource.

Visit the archive at www.arbeiter-zeitung.at

Don’t miss our free Table Models webinar today! You will learn how to transform historical tables into searchable, structured text.

Register here: www.transkribus.org/event/table-...

If you already have experience in training custom AI models with Transkribus, take your text recognition model training to the next level with this advanced webinar. This webinar recording covers… Advanced Model Training in Transkribus (English)

Ready to improve your transcription results?
Our latest webinar recording is now available to watch on YouTube! Find out how to train custom AI models that can recognise unique historical handwritings.

How the University of Helsinki teaches Transkribus to students The University of Helsinki equips students with Transkribus skills, aiding historians in text recognition and archival research for a digital future.

Digital tools are now essential for modern historical research.

At the University of Helsinki, Ville-Pekka Kääriäinen is leading the way by teaching students how to use Transkribus to process historical documents.

Read more about integrating AI into the curriculum:

Image: Frontispiece, Antonio de Nebrija’s Introductiones latinae (c. 1550). Biblioteca Gonzalo de Berceo, via Wikimedia Commons (Public Domain).

#onthisday in 1492, Antonio de Nebrija is said to have presented the first grammar of a modern European language, Spanish, to Queen Isabella I.

This work remains a fascinating example of early language standardisation!

Struggling with complex document structures?

Join our free webinars on Table Models (20 Jan) and Field Models (27 Jan) and learn how to train Transkribus AI to recognise the specific layout of your documents.

Sign up here: www.transkribus.org/events

Image: Royal Fireworks and Illuminations at Whitehall (1749). Via Wikimedia Commons (Public Domain)

Throughout history, New Year hasn’t always been celebrated on 31 December. Depending on the time and place, it has often marked the spring or autumn equinox, or even 25 December.

However you celebrate the turning of the year today, we wish you a great start to the next one! 🎇

Images:
John Bull's new Cakes and Gingerbread (1799). British Museum, via Wikimedia Commons (Public Domain)

Recipe Book (17th-18th c.). Wellcome Collection, via Wikimedia Commons (CC BY 4.0)

Gingerbread has a long history, from ancient Egypt and China, where spiced honeybreads were used in ceremonies and daily life, to medieval Europe, where spiced cakes became popular among the wealthy.
Want to give history a taste? Try baking a 17th-century recipe yourself 👇

Images:
Joseph Mohr Autograph (1820/25). Melody: F.X. Gruber. Courtesy of Stille Nacht Gesellschaft, via Wikimedia Commons (public domain)

Historical St. Nikolaus Church, Oberndorf. Courtesy of Stille Nacht Museum Oberndorf, via Wikimedia Commons (CC-BY-SA-3.0-migrated)

#onthisday in 1818, "Silent Night", written by Franz Xaver Gruber, was first performed at St. Nicholas' Church in Oberndorf, Austria.

It has since become one of the most recorded Christmas songs in history and has been translated into around 300 languages. 🎄🎶

How to read Latin manuscripts with AI Unlock the secrets of Latin manuscripts with AI-powered transcription models, enhancing historical research and digital accessibility for historical documents.

Latin connects centuries of European history, but diverse scripts make manuscripts hard to read.
Here are five AI models that can be used with Transkribus to create accurate, digital versions of Latin documents!

👉 blog.transkribus.org/en/how-to-re...

About to keynote my first conference in Japan! Speaking about @transkribus.bsky.social and all things cooperative tech governance, and the importance of community driven tech - and community walled gardens in this AI moment.

How the German Archives for Diaries preserve personal history for future generations Discover how the German Archives for Diaries are preserving personal histories through digitisation and citizen science, providing valuable resources for researchers worldwide.

"Without volunteers, nothing works for us."
 
The German Archives for Diaries (DTA) is on a mission to preserve the thoughts of "ordinary people." As Markus Schaber notes, diaries offer the social and emotional context that state archives often miss.

👉 blog.transkribus.org/en/german-ar...

Press & Media

Planning a Transkribus workshop for students or colleagues?

Have a look at the free teaching resources on our Press & Media page to help introduce new users to the platform: info.transkribus.org/press-and-me...

Thanks for asking! Simply login and you can see an overview of all public Sites here: app.transkribus.org/sites?view=p...

The work doesn't end with document transcription. For many projects, the real goal is public access.

The Noord-Hollands Archief makes its transcribed records accessible and fully searchable through Transkribus Sites.

👉 Browse the collections: app.transkribus.org/en/sites/noo...

Image: First issue of The Observer (1791), sourced via The Guardian

👉 www.theguardian.com/media/organg...

First issue of The Observer (1791), sourced via The Guardian

#OnThisDay in 1791, The Observer was founded.

Now known as the world’s oldest Sunday paper, its early days were troubled: founder W.S. Bourne faced heavy debts, leading the paper to accept government subsidies for editorial influence.
A reminder to consider context when evaluating historical media!

📜 Catch up on everything the @materialwills.bsky.social team have been up to with this latest project update 👇

#EarlyModern #DigitalHumanities #History #CitizenScience

@uoearchhist.bsky.social @uniofexeterhass.bsky.social
@leverhulme.ac.uk @transkribus.bsky.social
@zooniverse.bsky.social

19 januari 2026: Werken aan digitale geletterdheid met historische bronnen In een professionele leergemeenschap van vier bijeenkomsten word je meegenomen in de achtergrond van AI en Large Language Modellen en ga je zelf aan de slag met Transkribus en authentieke bronnen e…

💻𝘏𝘰𝘦 𝘬𝘶𝘯 𝘫𝘦 #𝘈𝘐 𝘪𝘯𝘻𝘦𝘵𝘵𝘦𝘯 𝘰𝘮 𝘩𝘪𝘴𝘵𝘰𝘳𝘪𝘴𝘤𝘩𝘦 𝘩𝘢𝘯𝘥𝘴𝘤𝘩𝘳𝘪𝘧𝘵𝘦𝘯 𝘵𝘦 𝘭𝘦𝘻𝘦𝘯?

Denk mee op 19 januari over lessen mbt het AI model
@transkribus.bsky.social & digitale geletterdheid, bij
@unileiden.bsky.social

Voor wie (middelbaar) onderwijs, geschiedenis & technologie wil verbinden

@neerlandistiek.bsky.social

Melissa Terras on why cooperatives are the future of AI Discover why Melissa Terras believes cooperatives like READ-COOP represent the future of AI, fostering community-driven, sustainable technological innovation.

Can AI be ethical and transparent? At READ-COOP, the unique European cooperative behind Transkribus, we prioritise democratic control and purpose over profit.

Read Melissa Terras' (@melissaterras.bsky.social) thoughts on why this model is a global blueprint: 
blog.transkribus.org/en/melissa-t...

And don’t miss today’s free webinar on training your own text recognition model!

👉 Register here: www.transkribus.org/event/traini...

YouTube video by Transkribus Transkribus Document Editor Webinar (English)

Preparing Ground Truth for model training?

We've just uploaded the recording of our webinar on editing, correcting and structuring your documents , perfect for anyone getting their data ready for custom model training.

youtu.be/i4uiwtUByQ0