The Future of Information Extraction – Be Part of TUC 2024! ✨ Feb 15-16, In-Person and Online. Get your Ticket >>

Printed Latin and Greek (also German, English, Italian) 15th-19th century | PyLaia

Free Public AI Model for Handwritten Text Recognition with Transkribus

Printed Latin and Greek (also German, English, Italian) 15th-19th century | PyLaia

The “NOSCEMUS General Model” is tailored towards recognizing Latin prints from the early modern period. Although the model is designed to recognize Latin prints set in Antiqua-based typefaces, it is also capable of recognizing passages in Greek and passages set in (German) Fraktur.

In creating the Ground Truth the following transcription guidlines were followed:
– ligatures (e. g. Æ or æ, Œ or œ) and standard abbreviations (e.g. -que, -us, -tur, …mm…, …nn…) have been expanded
– long s (ſ) was transcribed as a normal s
– small caps were transcribed as majuscules
– special characters and diacritics (e. g. &, ë, ï or ę) were kept

The model was released by Stefan Zathammer and it is based on training data coming from the Digital Sourcebook of the NOSCEMUS project (

If you use the Noscemus model as a base model for your own model, or if your edition is based on a transcription made with the help of the Noscemus model, you are kindly requested to mention the Noscemus model.

The NOSCEMUS project ( has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 741374).

Model Overview

Noscemus GM 6
Noscemus Project (University of Innsbruck)
Model ID:
15th, 16th, 17th, 18th, 19th
German, Greek, Latin
Latin alphabet, Gothic Script, Greek alphabet, Fraktur
CER on validation set:
0.8 %
Simply upload a picture and test this model

By uploading an image, you accept our terms and conditions and our privacy policy

Noscemus GM 6 is freely available to everyone

Get started with Transkribus and use it for your own Material
You can use this model to automatically transcribe Print documents with Handwritten Text Recgnition in Transkribus. This model can be used in the Transkribus Expert Client as well as in Transkribus Lite.
This AI model was trained to automatically convert text from images of historical Latin alphabet, Gothic Script, Greek alphabet, Fraktur documents into editable and searchable text.