How to check if a PDF has real text or pictures
Check whether a PDF contains selectable text, scanned page images, or both, and learn what to do when its words cannot be copied or searched.
A PDF page can look like a normal document even when it is only a picture of one. The quickest check is to count the text a PDF can actually provide: if a page visibly contains words but has no extractable text, it may be a scan or image.
Use the PDF Text Counter to check a whole file or individual pages. Then compare the count with what you can select and search in a PDF viewer. This helps identify text-based, scanned, and mixed PDFs.
Check the PDF for selectable text
- Open the PDF Text Counter and open your PDF in the tool. The file is processed in your browser and is not uploaded.
- Check the total word or character count. A nonzero count means the PDF contains text that the tool can extract. It does not guarantee that every visible word is represented correctly.
- Look at the count by page, if you need to check particular pages. Compare pages with text to pages that appear to be scans. A zero count on a page with visible writing is a clue that its writing may be part of an image.
- Open the same PDF in a viewer and try to select a sentence with your cursor or finger. Copy it, then paste it into a text editor. Also try searching for a word you can see on the page.
Selection and search are useful cross-checks. A page where words can be selected, copied, and found by search almost certainly has a text layer. If the words cannot be selected and the text count is zero, the page is likely image-only. Neither check is perfect on its own, so use both when the result matters.
What the count tells you
The page has a normal text count
The PDF contains extractable text. You should usually be able to search for words and copy passages, although odd characters or broken spacing can still point to a font or text-encoding problem. A count shows that text exists, not that the document is accurate or easy to read.
The page shows words but the count is zero
The page may be a scan or a picture embedded in the PDF. The letters are visible to you, but the file may store them as pixels rather than characters. Try selecting and searching in the viewer to confirm. If neither works, treat the page as image-only unless you can run optical character recognition.
Some pages have text and others do not
This is a mixed PDF. For example, a document might contain digitally written pages followed by scanned receipts, or a text-based report might include a photographed form. Check page-level results rather than relying only on the total. A nonzero file total can hide pages with no extractable text.
The count is nonzero, but copying gives strange text
The PDF has a text layer, but it may not map the displayed letters to the right characters. Copying can produce scrambled symbols, missing spaces, or unrelated letters. A count alone cannot verify that copied text matches what you see. Check a few lines against the page, especially if you plan to search, quote, or reuse the text.
What to do with an image-only PDF
If you need to find words, copy a passage, or use the text with assistive technology, run OCR on the image-only pages. OCR, or optical character recognition, tries to turn letters in a picture into selectable text. Use an OCR feature in a PDF or scanning app that you trust, then save a new searchable copy.
OCR can misread small print, handwriting, low-contrast scans, columns, or unusual fonts. Review the result before relying on it. Compare names, dates, amounts, and other details with the original page, since a scan can look clear while the recognized text contains errors.
If you only need to read the PDF, you may not need OCR. If you need reliable searching or copying, OCR is the practical next step. For a file with both text and scans, run OCR only on pages that lack usable text, if the app lets you choose pages.
Why a visible page may have no searchable text
A scanner often saves each page as an image inside a PDF. That preserves the page's appearance, but the words are not stored as text characters. The document can look crisp and complete while search, copy, and text counting return nothing.
Some scanned PDFs already have a hidden OCR text layer. The page still looks like an image, but its words may be searchable and countable. Check by searching for a word and copying a short sentence. If the copied wording is inaccurate, the text layer may have been recognized poorly.
There are also less common cases where text exists but a tool cannot extract it cleanly, such as unusual character encoding or text drawn as shapes. If the PDF Text Counter shows zero but your viewer can select and copy readable words, the file is not simply a collection of pictures. The issue may be how that text is represented or exposed.
Use page-level counts when only part of a PDF needs to be searchable, and keep the original unchanged while you test OCR. That gives you a way to compare the recognized text with the source pages.
