About Word Counter
Word Counter tallies words, characters with and without spaces, sentences, paragraphs and lines as you type — with a live reading-time estimate based on a standard 200 wpm pace. It's tuned for real prose, not just whitespace-splitting: it handles hyphenated words, em-dashes, smart quotes, and trailing punctuation the way an English teacher would. Writers use it to hit essay word limits, content marketers to stay inside SEO meta-description lengths, and students to enforce assignment caps. The whole thing updates with zero perceptible lag, even on 50,000-word manuscripts.
- No uploads
- Browser-only
- Works offline
- 100% free
How it works
- 1
Paste or type your text
Drop in anything from a tweet to a novel chapter. The counter is debounced for performance but updates in well under 100 ms.
- 2
Read the live stats
Word count, character count (with and without spaces), sentences, paragraphs, lines and an estimated reading time appear in a sidebar that updates as you edit.
- 3
Hit a target
Set a goal word or character count and the progress bar fills toward it. Useful for academic essays, social media caption limits, and search-result snippet budgets.
Counting words is fuzzier than it looks
Everyone agrees on character count until they don't. The deceptively hard question is "what is one character?" To a human, an emoji is one character; to JavaScript's string.length it can be two (a surrogate pair), and a flag emoji or a skin-toned emoji can be four or more code units. The unit you actually mean is a grapheme — a user-perceived character — and counting graphemes correctly requires the Unicode text-segmentation rules described in UAX #29 at unicode.org. For plain English prose the distinction rarely matters; for anything with emoji, accents, or non-Latin scripts, character counts from different tools will legitimately disagree.
Word counting has its own ambiguities. Is "don't" one word or two? Is "state-of-the-art" one or four? Is a number a word? Most counters split on whitespace and punctuation, which is a reasonable convention but not the only one — which is why your word count and Microsoft Word's can differ by a few percent on the same text.
How reading time is estimated
Reading-time figures come from a simple formula: word count divided by an assumed reading speed. The widely used range for silent reading of English is roughly 200-250 words per minute for average adults; a 2019 meta-analysis by Marc Brysbaert reviewed the literature and put typical silent reading around 238 wpm for non-fiction. A 1,000-word article at 230 wpm is therefore a little over four minutes. It's an estimate, not a measurement: technical material with code or formulas reads far slower, and skimmers move far faster.
Treat the number as a rough signal for readers, not a promise. The value of showing it is psychological — "5 min read" helps people decide whether to start — not precision. If you need a different speed (children, second-language readers, dense legal text), mentally scale: halving the wpm roughly doubles the time.
Character limits worth memorizing
| Where | Practical limit | What happens past it |
|---|---|---|
| SEO title tag | ~50-60 characters | Google truncates the displayed title in search results |
| Meta description | ~150-160 characters | Snippet gets cut with an ellipsis |
| Tweet / X post | 280 characters | Hard limit — you cannot post past it |
| SMS segment | 160 GSM characters (70 if any emoji) | Splits into multiple billed messages |
| URL slug | Keep well under 75 | Long URLs get truncated in shares and look spammy |
Search-engine pixel widths vary, so character counts are guidance — a title of wide letters truncates sooner than one of narrow letters.
What a counter doesn't tell you
- Word count is not readability. 300 plain words can be far easier than 150 jargon-dense ones; the count says nothing about sentence length or clarity.
- Counts can differ from Word, Google Docs, or a CMS because of how each treats hyphens, numbers, and punctuation — none is "wrong," they use different rules.
- For SEO, the pixel width of a title matters more than its character count; treat the character figure as a fast proxy, not the final word.
- If your text is padded with invisible or zero-width characters from a rich-source paste, the character count will read high — run it through Text Clean first; to draft placeholder text to a target length, Lorem Ipsum.
Frequently asked questions about Word Counter
How are 'words' counted exactly?
We split on Unicode whitespace and ignore empty tokens. Hyphenated words like 'state-of-the-art' count as one word (matching Microsoft Word's behaviour). Numbers count as words. URLs count as one word each. This matches what most teachers, editors and ATS systems expect.
Is the reading time accurate?
It uses 200 words per minute, which is the median adult silent-reading pace. Technical or dense text reads slower (150 wpm); light fiction reads faster (250 wpm). The estimate is therefore a useful ballpark for blog posts and reports but shouldn't be taken as precise — adjust your own expectations if you know your audience reads faster or slower.
Does it count Twitter / X character limits correctly?
Twitter counts most characters as 1 weight, CJK characters as 2 weight, and shortens URLs to a fixed 23-character allowance regardless of actual length. Our basic character count is raw codepoints — for accurate Twitter weighting use a dedicated tweet counter. The character-without-spaces stat is closer to what SMS providers count.
Privacy, offline use, browser support, and pricing questions are answered on the site-wide FAQ.