PDF & Files
PDF Word Counter
Free to useNo sign-up requiredNo watermarkRuns in your browser
Translators, editors, students and anyone quoting per word need a word count, and PDF readers do not give you one. This tool extracts the text from every page of a PDF and counts words and characters - with and without spaces - along with pages, average words per page and an estimated reading time.
It uses the same counting rules as the site’s word counter, so a document gives the same figures whether you paste its text or drop the PDF. Pages with no extractable text, such as scanned pages, are listed separately rather than silently counted as zero. Your file is processed locally in your browser and is not uploaded.
How this tool works
Choose a PDF
Drop it on the box or click to browse. Protected PDFs ask for their password.
Wait for the pages to be read
Progress is shown page by page; long documents can be cancelled at any time.
Read the totals
Words, characters with and without spaces, pages, words per page and reading time.
Check pages without text
If some pages are scans, send the file to the OCR tool to read them, then count that text.
How it works
pdf.js, running in this page, extracts the text layer of each page in turn - the same text you would get by selecting and copying it. Runs of spaces are collapsed and line ends kept, then the text is counted with the word counter’s rules: a word is anything between whitespace, and characters include spaces and line breaks unless stated otherwise.
Reading time uses the site’s standard 200 words a minute, rounded up to whole minutes. The average words per page is taken over pages that actually contain text, so a few blank or scanned pages do not drag it down.
A page with no extractable characters at all is listed as having no text. That is usually a scan or a photo of a page, which needs OCR (optical character recognition) to be read.
Common use cases
- Quoting a translation, proofreading or transcription job priced per word.
- Checking an essay, thesis or report against a word limit when you only have the PDF.
- Estimating how long a report or policy takes to read.
- Comparing the length of two versions of a document.
Limitations worth knowing
- Scanned or image-only pages contain no text to count. They are listed; run them through OCR first.
- Text inside images, charts and diagrams is not counted.
- Headers, footers, page numbers, footnotes and table cells are part of the text layer and are counted.
- Multi-column layouts are counted correctly for words, but the order of the extracted text may not follow the columns.
- A word hyphenated across a line break counts as two words; some PDFs also split words with odd spacing.
- Counts can differ slightly from Microsoft Word, which has its own rules for numbers, dashes and punctuation.
Formula
Words
sum of words on every page (text split on whitespace)
Average words per page
total words ÷ pages that contain text
Reading time
ceil(words ÷ 200) minutes
Worked examples
A translation quote
A 24-page contract gives 8,412 words and 51,030 characters with spaces - about 43 minutes of reading at 200 words a minute. The quote is based on the word count.
A partly scanned report
A 40-page report counts 9,600 words on 36 pages; pages 37-40 are flagged as having no text. They are signed scans, and the OCR tool reads them.
Frequently asked questions
Is my PDF uploaded?
No. Your file is processed locally in your browser and is not uploaded. Text extraction and counting happen on your device.
Why is my word count zero?
The PDF almost certainly contains pictures of pages rather than text - a scan, or a document saved as images. Use the OCR tool, which reads the text out of the images in your browser, then count that text.
Why does it differ from Microsoft Word’s count?
Word and this tool split words slightly differently around numbers, dashes and symbols, and the PDF includes headers, footers and page numbers that Word may treat separately. Differences are usually well under 2%.
Are footnotes and headers counted?
Yes, if they are real text in the PDF. The tool counts the whole text layer of each page; it cannot tell a running header from body text.
Can it handle very long PDFs?
Yes. Pages are read one at a time with progress shown, and you can cancel. Files up to 100 MB are accepted; very long documents take longer on older phones.