What the Word Frequency Counter does
The Word Frequency Counter takes a block of text and one specific word, and tells you how many times that word appears. That's the entire job.
Paste a paragraph, type the into the search field, and the counter reports — say — 14. Type and and you get 7. Type quickly and you get 2. The number updates as you type, so you can test multiple candidate words without re-pasting the source text.
This is the small-and-fast version of a frequency tool. If you already know the word you care about, you don't need a full distribution chart — you just need the count, in under a second, with no setup. That's what this is for.
When you'll use it
The use cases all share one trait: you already know what you're looking for, and you want the answer fast.
- Edit detection — "am I overusing this?" You're editing a draft and you notice you've leaned on the word just a lot. Paste, type just, get the count. If it's 30 across 1,000 words, you're overusing it. Fix it and re-check.
- SEO keyword targeting. You're writing for the search term aluminum recycling and you want to know how many times each word appears in the draft. Count aluminum, count recycling, compare.
- Citation checks. The reviewer flagged that you use the word argues too often. Open your manuscript, count argues, confirm or refute.
- Brand or product name verification. Marketing copy is supposed to mention the product name at least five times. Paste, search for the name, confirm.
- Translation review. The English source uses therefore nine times. Check whether the Spanish translation uses por lo tanto a similar number of times.
- Compliance and legal review. A contract template should mention indemnify exactly twice. Quick check before sending.
- Teaching. Show a student that their essay uses I 47 times in 600 words. Sometimes the count is the lesson.
- Reading comprehension audits. A textbook chapter should reinforce a key term. Count it. If it appears once, the chapter is under-teaching the term.
Each of these is a five-second job. The friction this tool removes is the friction of opening a fuller text-analysis app, waiting for it to load, hunting through a frequency table for one word, and copying the number out. Type, read, done.
A worked example
Take the opening sentence of the U.S. Constitution:
We the People of the United States, in Order to form a more perfect Union, establish Justice, insure domestic Tranquility, provide for the common defence, promote the general Welfare, and secure the Blessings of Liberty to ourselves and our Posterity, do ordain and establish this Constitution for the United States of America.
Paste it into the counter and search for each of these:
| Search word | Count | Notes |
|---|---|---|
| the | 6 | The most common word in the passage, as usual for English |
| and | 3 | Conjunction-heavy prose |
| of | 3 | Tied with and |
| establish | 2 | Repetition is intentional — bookends the sentence |
| United States | 0 | This is two words — the counter searches single words |
| states | 2 | Counting states as a single word gives the right answer |
| justice | 1 | Case doesn't matter — capitalized in source, found anyway |
Two things to notice. First, the counter is case-insensitive by default, so Justice in the source matches a search for justice. Second, the search is for individual words, not phrases — searching for United States returns 0, but searching for states alone correctly returns 2. For multi-word phrases, you'd need a phrase search (which most browsers' built-in Find/Cmd-F handles fine).
Case sensitivity, exact match, and the obvious traps
The counter has two switches that catch most edge cases:
- Case sensitivity (off by default). With case insensitivity on, Cat, cat, and CAT all count as the same word. Turn case sensitivity on if you need to distinguish polish (verb) from Polish (nationality), or apple (fruit) from Apple (company). This is the most useful switch in real editing work.
- Whole word only (on by default). With this on, searching for cat finds cat but not cats, catalog, or category. Turn it off if you want to match substrings — useful for finding all forms of a verb (run, runs, running all contain run) at the cost of false matches.
The default settings (case insensitive, whole-word match) are correct for about 90% of jobs. The two situations where you'll change them:
- You're searching for a proper noun and want to avoid catching a common-noun homograph → case sensitive, whole word.
- You're searching for a verb and want to catch all conjugations → case insensitive, partial match. Then mentally subtract any false positives (cat matching catalog).
Whole-word vs partial-word: the punctuation question. "Whole word" means the search term must be surrounded by word boundaries — whitespace, punctuation, or the start/end of the text. A search for cat with whole-word on matches cat, and cat. and (cat) but not cats or catfish. If your source text uses unusual punctuation — em-dashes, smart quotes, ZWJ characters — the boundaries still work because they're all non-word characters in Unicode.
How it counts
Everything runs in your browser using JavaScript. Nothing is uploaded.
The internal logic, simplified:
- Take the source text and the search term.
- If case sensitivity is off, lowercase both.
- Build a regex:
\b(search)\bfor whole-word, or just(search)for partial. - Count matches.
The \b is a word boundary — it matches the position between a word character (letter, digit, underscore) and a non-word character (space, punctuation, start/end of text). This is what makes whole-word search work correctly.
The counter is bounded by browser memory, not by the tool. Practical limit is millions of characters with no noticeable lag.
Counter vs Analyzer — when to use which
The Counter has a sibling: the Word Frequency Analyzer. The analyzer shows you the full distribution of every unique word in the text, ranked. The counter answers one question about one word.
| You want to know... | Tool | Why |
|---|---|---|
| "How many times does therefore appear?" | Counter | One word, one answer — faster |
| "What words am I overusing?" | Analyzer | You need the discovery, not a single answer |
| "Is my keyword density 1% or 3%?" | Either | Counter is faster if you know your total word count; analyzer gives % directly |
| "What's the top 10 most common words?" | Analyzer | Counter can't rank — it only knows about words you search for |
| "How often does my product name appear?" | Counter | You know the name; you want the count |
| "What's this text actually about?" | Analyzer | Discovery question, not measurement |
The rule of thumb: if you know the word, use the Counter. If you don't know what you're looking for, use the Analyzer.
The "am I overusing this word" workflow
This is the most common use of the Counter, and it deserves its own playbook. Editing your own writing is hard because the brain auto-completes phrases you've used before — you don't notice you've written really for the eighth time in 500 words.
The workflow:
- Finish a draft. Don't run this mid-writing; you'll over-edit.
- Run your draft through the Word Counter for the total word count. Note the number.
- Open the Word Frequency Counter and paste the draft.
- Search for the usual suspects: just, really, very, actually, basically, simply, that, like (for casual writing); therefore, however, moreover, furthermore, thus (for academic); important, key, crucial, essential, vital (for marketing).
- Calculate density. Divide each count by your total word count. Anything above 1% for a non-content word is a red flag.
- Cut and re-check. Edit, repaste, recount. Repeat until you're satisfied.
Most writers find that one or two words are doing a disproportionate amount of work. Cutting them — or rotating in synonyms — is the highest-return edit you can make in five minutes.
Tips and quirks
- Apostrophes are part of words. It's and its are different searches. Search each separately if you're worried about a specific contraction.
- Hyphenated compounds count as one word. A search for long with whole-word on won't match long-term. Either search the full hyphenated form or turn whole-word off.
- Multiple paragraphs are fine. The counter doesn't care about paragraph breaks; it counts across the entire pasted block.
- The percentage isn't shown directly. The Counter gives you the raw number. If you want density, run the source through the Word Counter first for the total, then divide.
- For multi-word phrases, use Cmd-F in your browser instead. Phrase counting is what the browser's built-in Find is good at; this tool is built for single words with proper word-boundary handling.
- Emoji and special characters are word characters too, in Unicode's sense, so a search for a specific emoji counts each occurrence correctly.
Why a single-word counter exists at all
You can do this with your browser's Find feature (Cmd-F or Ctrl-F). You can do it with a one-line script in any language. So why a dedicated tool?
Two reasons. First, browser Find counts substrings, not words — searching for cat with Cmd-F matches catalog, and there's no way to turn that off in most browsers. Second, browser Find shows you "1 of 14" but doesn't put the number front and center. The Word Frequency Counter shows you the number, large, on the page. Half a second faster, but you'll do this twenty times a week.
Big Software has built whole writing apps around the idea that they need to nag you about every problem in your text. This tool answers one question. That's the offer.
Related text tools
- Word Frequency Analyzer — the sibling tool. Shows every unique word in your text, ranked by frequency. Use it when you don't already know which word to investigate.
- Word Counter — total words, characters, sentences, reading time. The starting point for any text audit. Run it first to get your denominator for density calculations.
- Character Counter — character total, with and without spaces. For strict length limits.
- Readability Checker — Flesch reading ease, grade level, average sentence length. Pairs well with the frequency counter when you're auditing prose quality.
- Sentence Counter — sentence and paragraph counts. Useful when you want to know if your repetition problem is at the word level or the sentence-structure level.
- Text Repeater — the inverse job. Need to generate repeated text rather than measure it? That tool repeats a string N times with optional separator.
- Remove Duplicate Lines — for cleaning up word lists when you've exported frequency data and have duplicate entries to merge.
Frequently asked questions
Is the search case-sensitive?
No, not by default. A search for cat matches Cat, cat, and CAT. There's a toggle to turn case sensitivity on when you need to distinguish proper nouns from sentence-initial capitals — useful when searching for words like Polish, Apple, or May that are also common words in lowercase.
Can I count phrases instead of single words?
Not directly — the counter is built for single words with proper word-boundary handling. For multi-word phrases, your browser's built-in Find feature (Cmd-F on Mac, Ctrl-F on Windows) is the faster choice. It counts substring matches, which is what you want for phrases.
Does the counter match partial words?
By default, no. A search for run with whole-word matching on returns counts only for the exact word run, not runs, running, or runner. Turn whole-word matching off if you want to catch all forms of a verb at once — you'll need to mentally filter the false matches (run won't match around because around doesn't contain run, but a search for art would match start, article, and departure).
Is my text stored or uploaded?
No. The Word Frequency Counter runs entirely in your browser using JavaScript. The source text, the search term, and the count never reach any Microapp server. Close the tab and everything is gone.
What's the largest text I can paste in?
Practically, the limit is your browser's memory. The counter has been tested with documents of several hundred thousand words with no noticeable slowdown. For whole-book-sized inputs (a million words or more), expect a second or two of processing on first paste; everything after is instant.
Why doesn't searching for "the" find "There"?
Because There contains the as a prefix, not as a whole word. Whole-word matching (the default) requires the search term to be bounded by non-word characters on both sides. The r after the in There is a word character, so the match is rejected. Turn whole-word matching off if you want to catch substrings — searching for the would then match The, there, therefore, other, and so on.
Can I use this on text in languages other than English?
Yes for languages that use whitespace between words — Spanish, French, German, Portuguese, Russian, Greek, Arabic, Hebrew, and so on. Word-boundary detection works on Unicode letter classes, so accented characters and non-Latin scripts are handled correctly. For Chinese, Japanese, and Thai — which don't use spaces between words — word-boundary matching can't segment the text correctly; turn whole-word matching off and treat the result as a substring search.