Transkribus · Super Model · Published March 19, 2025
Dutch Dean (Super Model)
Description
Dutch Dean is a highly capable Transkribus Super Model for recognising handwritten and printed Dutch texts. Trained on an extensive Dutch-language dataset enriched with community contributions (special thanks to Pauline van den Heuvel from Amsterdam City Archives and Annemieke Romein, our community director, for coordinating a large part of community contributions), it is designed to handle a wide variety of documents, from medieval manuscripts to modern printed works. The model excels across different styles and periods of Dutch writing, offering reliable performance for historical research, archival work, and contemporary projects alike. Super Models are generalists and usually provide great out-of-the-box performance on a wide variety of materials. Users can then leverage the processing output of the Super Model for correcting and using it as training material for dedicated specialist models.
Try this model
Use this modelOpen in Transkribus
Very low error rate3.5% CER
Character Error Rate (CER) measures the percentage of characters incorrectly recognised. Lower is better. This model scored 3.5% on its validation set. As a rule of thumb, a CER below 10% is considered good for most handwritten material. Super Models are trained on very large and diverse datasets, making them robust across a wide range of handwriting styles and languages. Because of this diversity, a low CER on the validation set is a strong indicator of general-purpose quality.
Measured on the model's own validation data. Results on your documents may differ depending on handwriting style, document condition, language, and how closely your material resembles the training data.
Words24,721,830
Model ID309753