Unique Word Counter

How it is calculated

The tool finds every run of letters in your text, and a word may contain one straight apostrophe between letters (don't counts as one word). Digits, punctuation and symbols are not words. A hyphenated word is split at the hyphen, so well-known counts as two words.

Words are compared after being converted to lower case, so The and the are the same word. The results are total words, unique words and diversity = unique ÷ total × 100, shown as a percentage with two decimals (the type-token ratio). On the Turkish page, the dotted and dotless I are folded in the Turkish way (I becomes ı and İ becomes i). Text without any letters, or empty text, returns no result. There is no stemming, so run and running are different words.

Common mistakes

• Expecting numbers to count: in room 101 only room is a word, and a text made only of digits returns no result.
• Using curly apostrophes: a typographic ’ is not recognised inside a word, so don’t counts as two words, don and t, while the straight don't counts as one.
• Comparing diversity of texts with very different lengths: the ratio falls as a text grows because common words keep repeating, so a 50-word text and a 5,000-word text are not directly comparable.
• Assuming related forms count as one word: walk, walks and walked are three unique words because the tool does no stemming.
• Treating a hyphenated compound as one word: well-known is counted as well and known, which raises the total.