Number Extractor: Pull Figures Out of Prose
Runs 100% in your browser — your text never leaves this page, and nothing is uploaded or stored.
This number extractor pulls every number out of prose — integers, decimals or both — with a sum and average reported for the list. Use it to extract numbers from text and pull numbers from a document for spreadsheets and reports — integers, decimals or both, with a sum and average reported for the list.
How to use the Number Extractor
- Paste your document into the box — or press Load sample to mine the example order and invoice text.
- Choose the Number type: integers and decimals together, integers only, or decimals only.
- Tick Remove duplicates or Sort numerically if the raw in-order list is not what you need.
- Read the numbers in the result box, one per line, and check the summary note for the count, sum and average.
- Copy the list into your spreadsheet or calculator.
About this tool
Numbers embedded in prose — invoices, reports, logs, research papers — resist spreadsheets. This extractor lifts them all out in order, so you can paste them into Excel or a calculator, complete with an instant sum and average in the summary line. The extractor reads numbers the way documents write them: digits with optional thousands commas (1,250 is one number, its commas stripped), an optional decimal part, and a leading minus when one is attached — the −40 in a returns line keeps its sign. It is honest about ambiguity: a version string like 4.2.1 is not one number, so it is lifted as 4.2 and 1. The type filter narrows the haul. Integers only keeps whole numbers — a decimal is dropped whole, never truncated, so 4.2 does not sneak in as 4 — and Decimals only keeps the fractional ones. The summary note reports how many numbers were found in the text, plus the sum and average of the list after your filters and deduplication, rounded to two decimal places. Reach for it when an invoice needs auditing, a report’s figures need checking, or a log’s values need totalling. The extraction runs entirely in your browser, so financial documents and internal reports are never uploaded anywhere. For the addresses in the same text, the tool that can extract emails from text pulls them into the same kind of clean, counted list.Worked example
Here is the sample text that comes with the tool, extracted with the default settings (all numbers, in order): You pasteOrder 1042: 12 shirts at 19.99, 3 hats at 9.5 and 250 loyalty points. Invoice total 268.38 due in 30 days. Year 2026, version 4.2.1, rating 4.75 out of 5.You get
1042 12 19.99 3 9.5 250 268.38 30 2026 4.2 1 4.75 5.The note reads “13 numbers found · sum 3675.82 · average 282.76.” Two entries deserve a look: version 4.2.1 arrived as two numbers — 4.2 and 1 — because it is not a single number; and the list ends with “5.”, the rating sentence’s full stop hitching a ride on the 5. Switch the type to Integers only and every decimal, 19.99 included, drops out whole.
Features
- Extracts integers and decimals from any paste, in document order, one per line.
- Thousands separators merged (1,250 becomes 1250) and negative numbers keep their sign.
- Type filters: all numbers, integers only, or decimals only.
- A free summary with every result: how many were found, their sum and their average.
- Optional deduplication and true numeric sorting in the same pass.
- Private by design: extraction runs in your browser; your documents are never uploaded.