Troubleshooting guide
Paperless-ngx OCR Empty Text: When the PDF Imports but the Text Layer Is Blank
Category: Document Workflow
Troubleshooting
You have a watched import volume in Docker, the PDF appears in Paperless-ngx, but searching it returns nothing because the OCR text layer is empty. Start by confirming whether the file is actually image-only. Open the original PDF in a viewer and try to select text. If nothing highlights, the PDF has no embedded text and OCR is the only path. Next, check the consume log for that document. Run docker logs <container> 2>&1 | grep -i ocr and look for lines mentioning tesseract or OCR. A clean run with no errors but an empty result usually points to a missing or mismatched OCR language pack, or an image too low-resolution for tesseract to read. Confirm the language setting in your Docker environment, for example PAPERLESS_OCR_LANGUAGE=eng, and verify the corresponding tesseract data is present in the container. If the PDF is a scan at very low DPI or heavily skewed, OCR may return whitespace. As a controlled test, copy one suspect PDF into a separate folder, run pdftoppm -r 300 -png input.pdf page to rasterize it, then OCR a page with tesseract page-1.png stdout. If stdout is empty, the source image is the problem; if it contains text, the Paperless-ngx pipeline configuration is the problem. This is the practical paperless ngx ocr empty text diagnosis. After adjusting language or resolution, re-import a copy and search for a known word. If results are still blank, roll back by restoring the previous OCR language setting and re-running the consume step on the original file so you do not lose the document. For a guided review of your Docker OCR configuration,
