Count the Words and Characters in a PDF

Get a word count, character counts, pages and a reading-time estimate from a PDF in one step.

About the Word Counter for PDF

Word processors show a live count as you type, but a PDF gives you nothing — which is a problem when a submission has a hard limit, a quote depends on length, or you need to know how long a document takes to read. This tool reads the text of every page and reports six figures: total pages, words, characters excluding spaces, characters including spaces, the average words per page, and an estimated reading time.

Counting is done properly rather than by splitting on spaces. Words are matched against letters and digits from any alphabet, with apostrophes and hyphens treated as part of the word, so contractions and hyphenated compounds each count once and accented or non-Latin scripts are handled correctly rather than being miscounted. Nothing is produced or downloaded — the figures appear as a table. Since the text is read inside your browser, checking a confidential draft costs nothing in exposure.

Key Features

  • Six figures at once: Pages, words, both character counts, words per page and reading time all appear together.
  • Proper word matching: Hyphenated words and contractions count as one word, and non-Latin alphabets are counted correctly.
  • Both character totals: Counts with and without spaces are given, since different guidelines ask for different ones.
  • Reading time estimate: Length is converted into minutes at a standard reading pace, useful for scripts and presentations.
  • Nothing is modified: The document is measured and left alone — no file is written or altered.

How to Use the Word Counter for PDF

  1. Add the PDF you need to measure.
  2. Run the count and wait while each page's text is read.
  3. Read the table for the figure your guideline actually specifies.

How It Works

Text is pulled from every page with pdf.js, joined together, and matched against a Unicode-aware pattern covering letters and numbers plus apostrophes and hyphens, so the count is not thrown off by punctuation or by scripts without spaces between words in the Latin sense. Reading time divides the word total by a standard prose pace of a little over two hundred words a minute and rounds up to at least one. The limitation is the same one that affects any text-based PDF tool: a scanned document contains images of words rather than words, so there is no text layer to count and the totals will come back as zero. Headers, footers and page numbers are counted too, since they are part of the text layer.

When You Need It

A doctoral student assembling a thesis chapter has a hard limit set by the university, and the supervisor returned the draft as a PDF rather than the original document. Knowing the count before submission avoids a formatting rejection at the deadline. The chapter contains unpublished findings that are not yet through examination, and many institutions have explicit rules against putting unexamined work into third-party web services. Editors quoting on a manuscript and speakers timing a script measure documents for their own practical reasons.

FAQ

Frequently Asked Questions

Why does my scanned PDF report zero words?
A scan is a picture of a page with no text layer behind it. This tool counts the text a PDF actually stores, and there is none, so nothing can be totalled. Character recognition would be needed first.
Does the count include headers, footers and page numbers?
Yes. Everything in the text layer is counted, so running heads and numbering are included. Expect a modestly higher figure than a word processor would show for the same body text.
How is the reading time worked out?
By dividing the word count by roughly two hundred and twenty words per minute, a common average for adult reading of prose. Dense technical material realistically takes longer.
Which character count should I use?
Check the guideline you are working to — publishers and funders differ. The count without spaces is the more common requirement, which is why both are shown.
Is my draft transmitted in order to be counted?
No. The text is extracted and totalled inside your browser, so an unpublished chapter is measured without a copy existing anywhere else.