How to use
- Paste your text into the Text to analyze box. A sample is filled in to start.
- With the defaults (stop words removed, minimum length 2), the table shows each word’s count and share right away.
- Switch Group by to 2-word or 3-word phrases to see expressions that appear together.
- To leave out particular words, type them in Words to exclude, separated by commas.
- Save CSV opens cleanly in a spreadsheet, or Copy result puts the CSV on your clipboard. The table shows the top 100 rows; the CSV contains everything.
What each option does
| Option | Behavior |
|---|---|
| Group by | Counts single words, runs of 2, or runs of 3 consecutive words. Phrases are built from the words left after filtering, and sentence boundaries are ignored |
| Strip Korean particles | Removes one trailing Korean particle so 사과를 and 사과는 count as 사과. It has no effect on English text |
| Remove stop words | Drops very common words such as the, and, of (a short built-in English and Korean list) |
| Include numbers | Counts number-only items like 2026 or 1,000. Off by default |
| Case sensitive | When off, Cat and cat are the same word |
| Minimum length | Words shorter than this are skipped. Default 2 |
How the numbers are calculated
- Share (keyword density) = occurrences of the word ÷ total words × 100. The total is counted after stop words and short words are removed, so changing an option changes every percentage.
- Words with the same count are listed in the order they first appear.
- Words are split with the browser’s
Intl.Segmenter.
For the sample text (This tumbler keeps drinks hot for hours. The lid is sturdy and the tumbler looks great. I love this tumbler, and the lid never leaks.) the defaults give 15 words in total, 12 of them unique. tumbler appears 3 times (20.0%) and lid 2 times (13.3%). Turning stop words off raises the total to 24, and the climbs to 3 uses, level with tumbler.
The Korean sample (이 텀블러는 보온이 정말 좋아요. 텀블러 뚜껑도 튼튼하고 보온 시간이 길어요. 디자인도 예쁜 텀블러를 샀어요.) gives 14 words, 11 unique, with 텀블러 at 3 (21.4%). With particle stripping off, 텀블러는, 텀블러 and 텀블러를 are counted separately and all 14 words are unique.
About Korean particle stripping
Korean attaches particles to nouns, so 사과를 (apple + object marker) and 사과는 (apple + topic marker) would otherwise be counted as different words. This tool applies a simple rule to the end of each Hangul word. It is not a morphological analyzer. It removes the longest matching particle once, and only when enough of the word remains: a one-character particle is removed only from words of three or more characters, and a longer particle only if at least two characters are left.
| Input | Result | Note |
|---|---|---|
사과를 |
사과 |
Works as intended |
학교에서 |
학교 |
Two-character particle 에서 |
서울로부터 |
서울 |
Longest particle 로부터 wins |
책을 |
책을 |
Two-character words keep a trailing one-character particle |
고양이 |
고양 |
False positive: the final 이 of “cat” is mistaken for a particle |
텀블러입니다 |
텀블러입니다 |
Verb and copula endings are not handled |
If you need exact counts for Korean text, turn stripping off and read the table yourself. Stop words are checked before stripping too, so 하지만 (“but”) is not cut down to 하지.
Keyword density and repetition
Writers use word counts to spot repeated phrasing and to see how often a target keyword appears. Repeating a keyword unnaturally can read badly and may look like keyword stuffing to search engines, but there is no agreed percentage at which it becomes a problem. Treat the share as a rough guide and put natural writing first.
Related tools
Check the overall length with Character Count, find repeated lines with Remove Duplicate Lines, compare drafts with Text Diff, or swap phrases with Find and Replace. Your text is processed in this browser and never sent to a server.