Ovrin strongly prefers native text extraction when the PDF already contains a text layer. This is faster and more precise than OCR because it comes directly from the document content rather than from an image interpretation pass.
Why this matters
Text-layer extraction is high quality and cheap. It keeps the extraction pipeline accurate and reduces unnecessary OCR work.