0 requests since your file/works offlineYour desk
Fix · a common PDF problem

Scanned PDF not searchable?

Short answer. A scan is a picture of words, not words. Pages without a text layer can’t be searched or copied until OCR reads them and adds invisible text over the image.
Page checkRead in this tab, never run
Drop a PDF anywhere on this pageThe instant X-ray reads the raw bytes; the full inspector then parses every object in a Web Worker. Nothing is uploaded or executed.
What you’ll see
  • Every page’s size and saved rotation
  • Which pages have a text layer and which are only images
  • The fix for each: rotate, organize or OCR
Requests since you added a file: 0Counted by this page · check it without trusting us

Causes and fixes

  1. 1
    Pages are images with no text layerScanners and phone apps often save only the picture.
    OCR the pages
  2. 2
    Pages were flattened to imagesSome compressors and redaction tools turn pages into images, removing the text layer.
    OCR the pages
  3. 3
    Text exists but is unusualSome PDFs map characters oddly, so search finds gibberish.
    See the extracted text

What is going on

Search, copy and screen readers all work on a PDF’s text layer: the actual characters the file stores. A scanner creates an image of the page, and unless it also runs optical character recognition, there are no characters to find. The page check above shows, page by page, whether a text layer exists.

OCR looks at the image and recognises the words in it. Here it runs in your browser with Tesseract, an open-source engine, for English and Spanish. The recognised words are written as invisible text exactly over the words in the picture, so the page looks the same but becomes searchable. The language data is downloaded from this site the first time, then works offline.

OCR is not perfect. Accuracy depends on the scan’s resolution, contrast and font; we do not publish an accuracy figure because we have not measured one on a corpus. For anything that matters, such as amounts and account numbers, check the recognised text against the image.

Checked 2026-10-05. The diagnostic above runs in this tab; your file is not uploaded.