Word Counter and Character Counter
Words and characters, clearly counted. Paste your writing to see its length, reading time and the phrases you use most.
Your text
Private browser-local counting. Up to 200,000 UTF-16 units and 1 MiB UTF-8. Drafts are saved only when you choose to save them.
Files replace the editor only after validation. Markdown is treated as plain text.
Your counts
- Words
- Characters
- Without whitespace
- Sentences
- Paragraphs
- Reading time
Detailed statistics and English readability estimates
Character categories overlap: Unicode code points, graphemes and encoded units measure different things.
- Unicode code points
- Original UTF-16 units
- Original UTF-8 bytes
- Characters without line breaks
- Whitespace code points
- Letter code points
- Digit code points
- Punctuation code points
- Symbol code points
- Emoji sequences
- Unique graphemes
- Unique words
- Lines
- Estimated English syllables
- Publisher words (6 characters)
- Average word length
- Average words per sentence
- Average characters per sentence
- Shortest sentence (words)
- Longest sentence (words)
- Speaking time
- Handwriting time
- Estimated pages
- Flesch reading ease estimate
- Flesch–Kincaid grade estimate
- ARI estimate
Reading speeds and page estimates
These defaults are assumptions. English readability uses estimated syllables, not a pronunciation dictionary. Times round upward to a second; ratios show one decimal.
Writing goals and limits
No goal selected.
Goals never cut your text. These are writing budgets, not platform enforcement or SEO ranking rules.
Keyword density and repeated phrases
Top 20 phrases. Density uses all same-size phrase windows before filtering. Phrases never cross sentence or paragraph boundaries.
| Phrase | Occurrences | Density |
|---|
Keyword CSV contains words derived from your text. All rows are exported, not just the visible top 20.
Sentence rhythm and lengths
Short: 1–9 words; medium: 10–19; long: 20–29; very long: 30+. These labels describe length, not writing quality. First 100 sentences shown; totals include all.
| Sentence | Words | Length |
|---|
Find, replace and text cleanup
Case changes apply to the selection, or the whole text when nothing is selected. Simple title case is not an editorial style guide.
Removing invisible format controls can change appearance. ZWJ and ZWNJ are retained. Every cleanup is explicit and undoable.
Optional drafts on this device
Saving keeps your writing on this device. Other device users may access it. Nothing is saved by default. Export a text copy before deleting drafts.
Autosave off. No selected draft.
Five named drafts, up to 50,000 UTF-16 units each, five previous versions each, and 2 MiB total. Storage failures leave your editor text unchanged.
Writing session timer
Measures net word gain during active time. Editing existing text may reduce the gain; this is not a typing-test score. Nothing is saved.
Keep a copy
Statistics JSON contains scalar counts and assumptions, not your source or keywords. Text downloads and optional print text contain your writing.
Counting assumptions
How words and characters are counted
“Hello world.” has 2 words, 12 characters including its space, 11 without whitespace and 1 sentence. A combined accent or joined family emoji counts as one visible character. Technical code points, UTF-16 units and UTF-8 bytes are shown separately. CRLF and CR become LF for analysis only.
Word and sentence boundaries use the selected language's browser dictionary. Blank lines divide paragraphs; adjacent nonblank lines stay in one paragraph. A final sentence fragment counts. Frequency is case-insensitive with Unicode normalization. Sentence and readability estimates can differ across editors.
Reading seconds = words ÷ chosen WPM × 60. Pages = words ÷ chosen words per page. Publisher words = non-whitespace characters ÷ 6. English readability uses the disclosed vowel-group heuristic, Flesch ease, Flesch–Kincaid and ARI; it cannot evaluate meaning or writing quality.
Sources: Intl.Segmenter, QCA readability research, NIST readability formula reference. Reviewed 16 October 2026.
Questions about counting text
Why does my word count differ from another editor?
Editors use different rules for punctuation, compound words and languages without spaces. This counter uses word-like Intl.Segmenter segments in your selected language. Browser dictionary versions can differ. Characters are visible Unicode grapheme clusters, so a combined accent or joined emoji can be one character.
Is my text saved or sent anywhere?
Counting, files and exports run in your browser. Text is not sent to Vicaya or included in analytics or URLs. Draft storage is off by default. Saving or enabling autosave keeps text on this device, where other device users may access it. Delete saved drafts when finished.
Are reading time and readability exact?
No. Speeds and page sizes are editable assumptions. English syllables use a simple vowel-group heuristic that misses irregular words; grade scores are estimates, not reading ages or a judgement of quality. Non-English and unsupported word forms have no readability score. Samples under 100 words are especially unreliable.
Do character goals enforce social-media limits?
No. Goals are user-editable writing budgets. Services may count weighted URLs, emoji or encoded message units differently. A 160-character snippet is not a Google ranking limit, and a short-post goal does not guarantee platform acceptance.
Updated