Extract phone numbers

The real case

A list of contacts lands in an email, a group message or a filled-in form, and the numbers are scattered among the sentences. Copying them by hand is the fastest way to get a digit wrong. Here you paste the whole text and get one number per line, ready to drop into a spreadsheet or an address book.

How they're recognized

A phone number is recognized when there's a long enough run of digits, which may contain the international prefix (+39), spaces, dots, dashes, slashes and parentheses: those are the shapes people actually write numbers in, and they're all accepted. Below nine digits in total the run isn't picked up, otherwise every year and every amount would end up in the list.

Why you should always read it back

No program can know what a run of digits means without understanding the sentence around it, so numbers that aren't phones get picked up too: an order code like 2024-00123456, a date written 12.03.2024, an invoice number, a broken-up IBAN. They're false alarms that are easy to spot by eye, and it's worth skimming the list before you use it. The other way round, a landline written with few digits and no separators may be missed: those are the two possible errors, and it's better to know them.

Watch out for duplicates: if the same number appears twice in the text, it appears twice in the list. In a spreadsheet they're removed in a second, but it's better to know before you count how many contacts you have.

Nearby tools

If you also need the email addresses out of the same text there is Extract emails from text, and for links Extract links from text. If the contacts are in a file exported from a phone instead, the right page is Open a .vcf address book, which reads names, numbers and addresses and turns them into a table. For every number, not just phones, there is Extract numbers from text.