Free document tools
Count the words in your PDF.
See which pages they come from.
Get words, characters and a page-by-page breakdown before quoting, editing or translating a document. Your PDF stays in this browser.
Choose a PDF to count
Up to 50 MiB and 300 pages. Counts the existing text layer; image-only pages need OCR first.
File contents, names and counts are not sent to Scanfile. Anonymous tool-use events follow the site’s privacy settings.
How to count words in a PDF
01
Choose your file
Select a PDF from your device. Analysis happens locally without sending the document to a server.
02
Check the page breakdown
Review words and characters per page. Pages with no extractable words are highlighted so a scan is not mistaken for an empty document.
03
Save or use the result
Download a CSV count report, identify a page range, or open the PDF translator when you need another language.
Understand the count
Why does this differ from Microsoft Word?
PDFs store positioned text fragments. This tool joins nearby text runs and counts word-like segments using your browser’s Unicode word segmentation. Reading order, line-end hyphenation, ligatures, headers and footers can change the result. Characters are normalized Unicode code points, including reconstructed whitespace. A simpler Unicode letter-and-number counter is used on browsers without word segmentation.
Can it count words in a scanned PDF?
Only if the PDF already has a usable OCR text layer. This tool does not recognize words in pictures. A zero-word page can also be blank or contain diagrams; it is not a definitive scan detector.
Can it count Chinese, Japanese or Arabic text?
The browser’s Unicode segmentation can identify word-like segments in many scripts. Word boundaries differ by language and browser, so use character counts and a manual spot check when comparing multilingual documents. This tool does not identify the document language.
Is my PDF uploaded or saved?
The counter reads the file in browser memory. It does not upload the PDF, save its contents to browser storage, or send file names or counts to analytics. Choosing the separate translation service is a new action with its own upload and privacy policy.
Can I use the number to price a translation?
It can help estimate the amount of extractable text. Confirm the counting rules with your translator, especially for scans, tables, repeated headers and languages without spaces. Scanfile PDF translation uses page-based pricing; this word count is not a price estimate.
Why is analysis limited?
This free tool processes up to 50 MiB, 300 pages and two million extracted text characters, with a two-minute timeout. Those bounds protect browser responsiveness. Use a smaller page range for larger or unusually complex documents.