Skip to main content

OCR Models

The Tesseract engine used by the OCR node is built in and free — it requires no database configuration, its settings are set directly on the pipeline node.

To use an AI-model-based OCR engine (for example Mistral OCR), register it from Manager → Settings → AI Models (OCR Models section):

  • Name — label shown in the OCR node's selector.
  • Provider — chosen among the OCR-type providers configured by the administrator (Mistral as of today).
  • Model name — e.g. mistral-ocr-latest.

A "Test" button generates a test image and runs a real OCR call with the entered configuration, to validate it before using it in production. Once registered, the model becomes selectable in the pipeline editor's OCR node ("Configured OCR model" option).

If no OCR-type provider appears in the selector, ask an administrator to configure one from Providers.