Appearance
Transkribus Developer Platform
This is the Transkribus Developer Platform: the developer documentation for using Transkribus programmatically, and the home of its APIs as they roll out. If you are new, read How the platform works first; if you want to run something now, Your first job gets you to a first result.
Beta: the Transkribus Developer Platform is in beta. The API Reference always reflects the current state of the API: every operation listed there exists and works. New API surface is announced in the changelog.
The first API surface turns images of (historical) documents into structured text using Transkribus text recognition technology. You submit a single image, as a base64-encoded string or an http(s) URL, in a POST request, and each job returns a processId used to track it and retrieve the result as JSON, a ZIP archive, or PAGE XML.
At a glance
| Access | Transkribus account on the Scholar plan or above |
| Credits | API processing costs 50% of the same job in the Transkribus platform |
| Results | Retained for 24 hours after completion |
| Images | JPEG, PNG or TIFF up to 20 MB, as base64 or URL |
Where to go next
- Authentication: get an access token with your Transkribus credentials
- Your first job: submit, poll, download in three requests
- How jobs work: the shared job contract
- Files and documents: getting material in and results out
- Models: what jobs run with, and how to choose one
- API Reference: every endpoint, schema and example