Skip to Main Content

Research Data Services

Background and links to more information about data management issues.

Getting Started with Transkribus

Transkribus is an advanced platform for the transcription, recognition, and analysis of historical documents. By leveraging machine learning and artificial intelligence, it transforms handwritten and printed text into digital format, facilitating easier access, search, and analysis of historical texts.

Currently, researchers face challenges in processing historical documents and transcriptions due to the time-consuming and error-prone nature of manual transcription. This limits access to valuable collections and hampers research efficiency. Transkribus addresses these challenges by providing an efficient, accurate, and versatile tool that aligns with our goals of improving research support and by preserving and expanding access to our historical document collections.

Dartmouth Libraries has an organizational (Epoch plan level) subscription which enables up to 25 "user seats"  with a total of 1 TB of storage and 60,000 credits per year (1 credit ~ 1 page of handwritten text + lines.  API uses 0.5 credit per page). The plan includes standard and advanced AI recognition and training functionality for text and layout, including: full text search, transcription editor, collaboration tools, API access, advanced data export formats.

Interested in trying out Transkribus? Register for an individual account to process up to 100 pages per month and then contact ResearchDataHelp@groups.dartmouth.edu to get added to the organizational plan.  We can assist with scoping out your project and working with Transkribus.

Search the Transkribus help center for more information: