How to use
- Paste your text, or click Load sample to try it out.
- Choose words or phrases, and set the minimum length and what to ignore.
- Review the ranking and click Download CSV to export everything.
Options explained
- Tokenization
- Uses the browser’s built-in word segmentation, which also handles Chinese, Japanese and mixed-language text.
- Phrases
- Counts adjacent two- or three-word combinations, such as machine learning, without crossing sentence boundaries.
- Stop words
- Common function words (the, of, and, is…) are skipped so the meaningful words rise to the top. Turn this off to count everything.
Worked example
Take this text:
The cat sat on the mat. The cat ate the fish. Machine learning helps machine learning researchers; the cat sleeps.The summary reads Words: 20 and unique: 10, because stop words such as the and on are counted as words but then skipped. The top rows with Count by set to Words:
| Word | Count | Share |
|---|---|---|
| cat | 3 | 21.43% |
| learning | 2 | 14.29% |
| machine | 2 | 14.29% |
Share divides by the 14 words that were counted, not the 20 in the text. With Two-word phrases, machine learning appears twice out of 8 phrases (25%). Phrases never cross a sentence or comma boundary, and a phrase that starts or ends with a stop word is skipped.
The page lists the top 100 rows; Show all lists the rest. Download CSV has every row with the columns term, count and share.
Frequently asked questions
How is this different from keyword density?
Keyword density is the share column — a word’s count divided by all counted words. Use it to see whether a term is overused in a page or essay.
How much text can it handle?
Hundreds of thousands of words run smoothly. Split very long documents into parts.
Is my text uploaded?
No. Everything is counted in your browser.
Privacy
This tool runs entirely in your browser. Your files and text are processed on your device and are never uploaded to our servers.