Extract phone numbers from a PDF
The numbers written in the document, as a list
Where it runs: On your machine.
Settings
For numbers written without a country code. Numbers that carry one are read in their own country.
Only the document's text is read, not its images. Each number comes out once, in international format, with its pages.
Drop a file to begin.
Everything happens in this tab. The file is never sent, so there is no upload, no queue and no size ceiling beyond your device's memory.
What this tool does
Reads the document's text and picks out every phone number, recognised against its country's numbering plan: a company number, a date or an IBAN is not mistaken for one. Each number comes out once, in international format, with the pages where it appears.
In the same family
Where the file goes
How to do it
- 01Drop in the PDF.
- 02Choose the country for numbers written without a country code, then start.
- 03Copy the list, or take it as a CSV for a spreadsheet.
Frequently asked questions
- How is a number recognised?
- Against each country's numbering plan, not by counting digits. A number is kept only if it could actually be assigned: right length, right prefix, a range that exists. That is what rules out company numbers, dates, case references and fragments of an IBAN.
- What is the chosen country for?
- For numbers written without a country code: “020 7946 0958” only means something in the United Kingdom. Numbers that carry one, written “+33”, “0044” or run into the number as in operator records, are read in their own country.
- What about a scanned document?
- A scan has no text layer: run it through “OCR” first, then come back. The numbers are then picked out of the recognised text, with one caveat: a misread digit gives a wrong number, or a discarded one.