How to use Word Counter
- Paste text or choose a .txt, .md, or .html file from your device.
- Read the live word, character, structure, byte, and time metrics.
- Revise the source as needed, then copy the figures you need for your workflow.
Supported behavior
Word Counter measures a document beyond a simple total. It reports word-like segments, characters with and without Unicode whitespace, grapheme clusters, code points, sentences, paragraphs, lines, UTF-8 bytes, and estimated reading and speaking times. Words are identified with the browser’s language-aware segmenter when available, so CJK writing and apostrophes are handled more naturally than with a basic space split; a Unicode-aware fallback covers older browsers. Hyphenated and apostrophe words normally remain single words, while standalone emoji do not become words. Use the separate character and grapheme figures when visual characters matter, because an emoji can contain several code points. Text and selected .txt, .md, or .html files are counted entirely in your browser: their contents and calculated statistics are never uploaded. The metrics are estimates where language boundaries are ambiguous, but they are useful for checking editorial limits, document size, and likely delivery time.
- Sentence and word boundaries remain estimates for abbreviations and languages not fully covered by the browser segmenter.
- HTML files are counted as their source text; markup is not interpreted as a rendered page.
Examples
Don't stop now.
Result: 3 words; 15 characters including spaces; 1 sentence
First paragraph.
Second paragraph!
Result: 4 words; 2 sentences; 2 paragraphs; 3 lines
Common use cases
- Check whether an article fits a publisher’s word and character limits.
- Estimate reading and speaking length before publishing a script.
- Inspect Unicode and UTF-8 size for localized interface copy.
Frequently asked questions
Does “don’t” count as one word?
Yes. Apostrophe words such as “don’t” normally form one word-like segment.
How are Chinese or Japanese words counted?
The tool uses Intl.Segmenter when the browser provides it, which offers language-aware CJK boundaries; older-browser fallback results can differ.
Why do graphemes and code points differ?
One visible character can combine several code points, as happens with family emoji and letters with combining accents.
Is HTML rendered before counting?
No. A selected HTML file is read locally and its source characters, including markup, are counted.