Which words are you actually using the most?
Paste a draft and you’ll see the answer in a ranked table: every word, how many times it shows up, and what share of the text it eats up. Top of the list first. The counting is case-insensitive, so “The” and “the” land in the same bucket, and punctuation gets stripped before anything is tallied.
Here’s where it gets useful. Writers repeat themselves without noticing. You think you said “leverage” twice. The table says eleven. That kind of thing is invisible when you’re reading your own prose, because your brain skips right over it.
How to read the results
Four columns, nothing fancy. Rank, the word itself, a raw count, and a percentage of the total counted words. The percentage matters more than people expect: if a single non-trivial word is pulling 4% of your entire article, that’s a red flag for repetition (or keyword stuffing, if you’re writing for search).
Two numbers sit above the table. Total words counted is everything that survived your filters. Unique words tells you how varied the vocabulary is. A 600-word post with 280 unique words reads richer than one with 140.
The two filters you’ll reach for
- Ignore common stop words. Flip this on and the usual suspects (“the”, “a”, “and”, “of”, “to”, and roughly 35 others) drop out. Without it, “the” wins basically every English document, which tells you nothing. Turn it on and the words that actually carry meaning float to the top.
- Minimum word length. Set it to 4 and you skip “is”, “it”, “an”, and other short filler. Handy when you only care about the substantive terms. Default is 1, so everything counts until you decide otherwise.
Both update the table live. No apply button, no reload. Change a number and the ranking reshuffles instantly.
Where this earns its keep
SEO folks use it to sanity-check density. Google doesn’t publish a magic ratio, but if your target keyword shows up at 3-4% across the page, that reads as forced. Aim closer to 1-2% and let the supporting terms do their work. The percentage column makes this a five-second check instead of a manual tally.
Editors run pasted chapters through it to catch crutch words. Every writer has them: “really”, “actually”, “basically”, “just”. Seeing “just” appear 40 times in a short story is a fast, slightly humbling way to start a revision pass.
Students and researchers lean on it for content analysis. Drop in interview transcripts or survey responses and the frequency table surfaces recurring themes without you reading every line twice. Then export to CSV and the whole ranking drops straight into a spreadsheet for charts or further sorting.
Translators and ESL teachers use it too, pulling the most common terms out of a passage to build a focused vocabulary list before a lesson.
Common questions
Does capitalization change the count?
Nope. Everything’s lowercased first, so “Apple”, “APPLE”, and “apple” all count as the same word. If you need case-sensitive counting, this isn’t the right tool for that.
How are hyphenated and apostrophe words handled?
Punctuation gets stripped, so “well-known” splits into “well” and “known”, and “don’t” becomes “dont”. It’s a deliberate trade-off that keeps the splitting predictable across messy real-world text.
What counts toward the percentage?
Only the words that pass your filters. If you’ve enabled stop words and set a minimum length of 4, the percentages are calculated against that filtered total, not the original word count. That way the numbers always add up to 100%.
Can I get the full list into Excel?
Yep. The Download CSV button gives you rank, word, count, and percent in a clean comma-separated file that opens directly in Excel, Google Sheets, or Numbers. Copy CSV does the same thing to your clipboard if you’d rather paste.
Is my text sent anywhere?
No. All the counting runs in your browser with JavaScript. Nothing gets uploaded, logged, or stored, so it’s fine for confidential drafts and unpublished work.
Will it work on languages other than English?
The counting works on any language that separates words with spaces, including accented Latin scripts and Cyrillic. The built-in stop word filter only knows English, though, so leave that toggle off for other languages. For full-document metrics, the Word Counter pairs nicely, and the Character Counter breaks things down letter by letter.