How do I count words in a PDF?
Choose a text-based PDF or try the demo. Review the word total and any uncounted pages. Open Choose pages to remove whole pages from your scope, then copy, download or print the report.
Check the text behind the count
Show the source preview next to the editable text. Find and replace, selected-text counts and review hints help inspect extraction. Corrections never edit the PDF. Original counts remain separate; enable reviewed counts to use corrected text in totals and reports.
Know what is being counted
The text-layer method counts Unicode words using the same segmentation as Taskono Word Counter. Hyphenated words and words split across lines count separately. Common ligatures expand before counting. Page totals sum to the scoped total. Review the original PDF when a count matters for an assignment or quote.
A PDF counting example you can try
The built-in demo contains two text pages and an image-only page. Choose Try a demo PDF above to reproduce these results.
| Page | Words |
|---|---|
| Page 1 | 22 |
| Page 2 | 15 |
| Page 3 | Uncounted: no text layer |
- Entire PDF: known words
- 37
- Pages with extractable text
- 2 / 3
- Include pages
- 1-3
- Exclude page
- 2
- Scoped known words
- 22
Page 3 remains uncounted in both scopes. These totals describe extractable text only, not a confirmed word count for the image-only page.
PDF extraction limits to check
- Tables and columns
- Text follows PDF content-stream order. Cells or columns can be interleaved, and their spacing may differ from the visual page. Check the page preview against the original.
- Ligatures
- Compatibility normalization expands common ligatures such as fi into fi. Unusual or broken font mappings can still lose text and affect counts.
- Hyphens and line breaks
- Hyphens split words. A word broken across lines is not automatically rejoined because the intended spelling cannot always be inferred.
- Scans and duplicate text layers
- Scans without text layers are uncounted. A scan with an OCR layer counts that existing text, including its errors. Duplicate or invisible text layers can inflate totals; the tool does not deduplicate them.
PDF word count questions
Can I count words in a scanned PDF?
Yes. Native text is extracted first; missing or unreadable pages are flagged. Choose those pages in local recognition, select English or Spanish, adjust rotation or contrast if needed and recognize them. Review and correct OCR text before relying on the count. Handwriting and exact reading order are not guaranteed.
Why does the count differ from Microsoft Word?
Word and PDF text extraction can include different document regions and use different word boundaries. Columns, tables, hyphens, ligatures and duplicate text layers can also affect PDF counts. Taskono does not promise Microsoft Word parity. Use the method required by your client or publisher.
Can I exclude references or selected pages?
Include or exclude whole pages with ranges. Selecting text in the review editor also shows its word and character counts. Automatic header or reference detection is not provided; verify document scope yourself.
Are PDF files uploaded to TASKONO?
No. PDF extraction and OCR run in browser workers. Recognition downloads engine and chosen language assets from Taskono only when requested; files, filenames and text are not uploaded or saved. Only estimate preferences can be saved on opt-in. Browser spellcheck starts off and may share text with its provider if enabled. Export work before reset or close.
Counting a pasted passage instead? Open Word Counter