DT 3b7 Server - OCR not working as expected

I only noticed this recently so perhaps it crept in in beta 7 of DT3 Server…but now when I drag an image-only PDF into DT3, it does not automatically create the PDF+Text version as it once did. If I highlight a document in my DT3 database and click OCR in the menu bar, it does do the OCR on the document and creates an OCR’d copy.

I have “Convert incoming scans:” set to “To Searchable PDF” in preferences.

By default DEVONthink doesn’t OCR images, only incoming scans. Did you use a smart rule to perform this? Another possibility is File > Import > Images… (with OCR)

Thanks. I guess I thought it automagically OCR’d incoming PDFs if they didn’t have a text layer already. The file that was failing was an image-only PDF (in other words, a scanned but not OCR’d document).

You could use a smart rule to perform this: