How the Word Frequency Analyzer Works
AnyTool Word Frequency Analyzer counts tokens entirely in the browser. It tokenises words with a Unicode-aware regular expression (runs of letters and digits, keeping intra-word apostrophes and hyphens), optionally folds case so The and the merge, tallies each distinct word, and ranks the results with a percentage share of all counted words. Phrase mode generates n-grams within sentence boundaries, and character mode counts individual characters by the chosen set.
- Words are Unicode letter/number runs, so non-English text is counted correctly
- N-grams (1–3 words) are built within sentence boundaries, never across a full stop
- Percentages are each token’s share of all counted tokens after filtering
- A type-token ratio (unique ÷ counted) summarises lexical variety
- All computation is client-side JavaScript — no server round-trip
