Extract Numbers From Text

Whole numbers, decimals such as 2.5, negatives such as -7, and comma grouped numbers such as 1,000 or 1,234,567. Each of the three categories has its own switch, so you can disable any format your data should not include.

It totals every number found after commas are stripped and any minus sign is applied, so a mix of positive and negative values adds up correctly. The sum uses standard floating point, so long decimal chains can show tiny rounding artefacts.

No. A digit run glued to letters, such as abc123 or 12px, is skipped so you only get standalone numbers. A word character directly before or after the digits blocks the match.

Only the leading number is taken, here 1.2, and the trailing .3 is ignored, so version strings and dotted identifiers do not spray out false matches.

They are listed in the order they appear in your text, top to bottom. There is no sort option, so if you need them ordered, copy the output into a number or line sorter afterwards.

A leading minus only counts as negative when nothing word-like sits before it, so 5-3 is read as 5 and 3, not as 5 and -3. Turn off Allow negatives to ignore minus signs entirely.

It ignores symbols like $, £ and % and reads only the numeric part next to them, so $19.99 yields 19.99. It does not parse scientific notation such as 1e6, which would be read as the number 1.

There is no fixed limit. The scan and the sum run in your browser, so the ceiling is your device memory, and even long documents are handled almost instantly.

Yes. Everything runs locally in your browser and nothing is uploaded to a server, so financial figures and reports never leave your device.