Skip to content
TextLab

Free word frequency counter

Word frequency counter for every word you used

TextLab's word frequency counter ranks every distinct word in whatever you paste by how many times it appears, with its share of the text beside it. Free to use, with no account and no ceiling on length: a full-length novel is tallied in about 26 milliseconds. The 127 commonest function words are hidden so the list says something, and you can set a minimum word length, a minimum number of uses, and whether The and the count as one word or two.

  • 100% free
  • No signup
  • Full ranked list
  • CSV download
  • Function words optional
0words in the listout of 0 different words in 0 of running text
letters
times
Order by

0
Running words
0
Different words
0
Used exactly once
0
Hidden by your filters

Every word you use will be listed here with its count and a bar showing how much of the text it occupies. Function words are hidden to begin with; untick the box to put them back.

How to count word frequency

Step two is what turns a raw tally into something you can act on.

  1. Paste the whole text, not a sample

    Frequency lists get more useful as they get longer, so give it the full chapter, the complete transcript or every survey answer at once rather than a paragraph. There is no upload step and no cap: a 144,000-word manuscript is tallied in about a fortieth of a second, and the list rebuilds itself while you keep typing.

  2. Tune the list until it says something

    The 127 commonest function words are hidden to start with, because otherwise the, of and and take the top three places in every text ever written. Raise the minimum length to four or five letters to strip out the remaining small change, raise the minimum count to two to see only what actually repeats, and untick the case box when you want Apple and apple counted apart.

  3. Reorder it or take it away

    Sort by count to find your tics, alphabetically to check a term is spelled the same way throughout, or longest-first to see the vocabulary you are asking readers to carry. Copy the list puts three comma-separated columns on your clipboard and Download CSV writes them to a file that opens straight into a spreadsheet.

Technical specifications

What gets listedEvery distinct word in the text, with its count and its share of the total running words, ranked highest first
Function-word filter127 articles, pronouns, prepositions, conjunctions and auxiliaries hidden by default, switchable in one click
Filters you setMinimum word length from 1 to 20 letters, and minimum number of uses from 1 to 999
Case handlingFolded by default, so The and the share a row; unfold it and capitalised forms are counted separately
Sort ordersBy number of uses, alphabetically A to Z, or longest word first
ExportCSV download or clipboard copy, three columns: word, times used, share of text to three decimal places
Measured speedAbout 1 ms for a 5,000-word article and 26 ms for 144,000 words, recounted on every keystroke
PriceFree, no signup, no row limit on the export

Frequently asked questions

Why are the, of and and hidden by default?

Because they win every time and tell you nothing. In ordinary English prose the single word the accounts for around six or seven per cent of everything written, and the hundred commonest words together make up roughly half of any running text — so a frequency list that includes them is the same list for a cookbook and a court judgment. The 127 function words hidden here are articles, pronouns, prepositions, conjunctions and auxiliary verbs; untick the box and they come straight back, which is worth doing if you are studying style rather than subject matter.

Are singular and plural counted as the same word?

No — cat and cats are separate entries, as are run, runs and running. Collapsing them means guessing at word families, and the guesses go wrong in ways you would not notice: the standard algorithms turn universities into universiti and reduce both meeting and meet to one entry, which quietly hides the thing you were looking for. Counting the exact forms is honest about what is on the page, and if you want the family total you can add the rows up yourself in the spreadsheet.

Does it count don't as one word or two?

One. An apostrophe inside a word does not split it, so don't, it's and o'clock each count once, and so do hyphenated compounds such as state-of-the-art. Numbers written with separators stay whole too, so 1,500 is a single entry rather than two. This is the same definition of a word that the rest of the site uses, so a total here always matches the one the word counter shows for the same text.

Why does my count differ from Ctrl+F in Word?

Because Find matches strings, not words. Searching for art in a document turns up article, artist, apart and heart, and Word's counter reports every one of them; this page counts whole words only, so art appears in the list exactly as often as somebody wrote the word art. The other common difference is case: Word's Find is case-insensitive by default and will merge The and the, which is what the case box above controls.

What is a word frequency list actually useful for?

Finding the thing you cannot see in your own writing. Editors run one over a manuscript to catch the crutch word that turns up eleven times in a chapter; teachers use one to show a student that their essay leans on three verbs; translators check that a term of art has been rendered the same way throughout; researchers coding open-ended survey answers start here before they start reading. In each case the value is the same — a count is not affected by how familiar the text has become to you.

How much text can it handle at once?

There is no limit, and the practical ceiling is your clipboard rather than this page. A 5,000-word article is tallied in about a millisecond; 144,000 words, roughly a full-length novel, takes around 26 milliseconds and still updates as you type. The full ranked list is rendered forty rows at a time with a button to show the rest, because a list of 12,000 distinct words is slow to draw, not slow to count.

Can I get the results into Excel or Google Sheets?

Yes — Download CSV writes a three-column file of word, times used and share of text, which both open directly. Copy the list puts the same three columns on the clipboard, so you can paste into a sheet without a download at all. The share column is written with a per-cent sign, so change the column format to text if your spreadsheet insists on reinterpreting it.

Why a frequency list always looks the same shape

Count the words in anything long enough and you get the same curve. A handful of words appear thousands of times, a few hundred appear often, and a very long tail appears once each. George Kingsley Zipf described the pattern in the 1930s: rank the words by frequency and the second-commonest turns up about half as often as the first, the third about a third as often, and so on down. It holds for novels, for legislation, for Twitter and for languages Zipf never studied. The practical consequence sits at the top of the list above — those first few rows are grammar, not subject, which is why hiding them is the default and why the interesting part of any frequency list starts around rank twenty.

The long tail is just as informative. In a text of a few thousand words, something like half of the distinct vocabulary will appear exactly once, and that proportion barely moves however much more you write — new words keep arriving at almost the rate old ones get reused. Counting words this way is one of the oldest computing tasks there is: Roberto Busa began indexing every word of Thomas Aquinas on punched cards in 1949, a project that ran for three decades and effectively invented the field. What he was building — a list of every word with every place it occurs — is what the table above gives you in a keystroke, minus the punched cards.

What you do with the list depends on why you counted. If you are editing, sort by count and look for the content word appearing far more often than its neighbours: that is usually a crutch rather than a theme. If you are checking a translation or a style guide, sort alphabetically and look for two spellings of the same term sitting next to each other. If you are working on a page that has to rank for something, the same counts expressed as percentages against a target phrase are what the keyword density checker shows, and it adds the two- and three-word phrases this page deliberately leaves out. And if the question is how wide your vocabulary is rather than which words dominate it, the unique word counter measures the shape of the tail instead of the head.

Counting happens where the text already is

The tally, the filters, the sorting and the CSV are all built by JavaScript running in this tab, so a manuscript, a set of confidential interview transcripts or a client's unpublished copy is counted without ever being sent anywhere. The download is generated in the browser too — the file never touches a server on its way to your Downloads folder — and nothing about your text or your filter settings survives a page reload.