Word & Character Counter
Count words, characters, sentences, and paragraphs as you type — with a reading-time estimate. Runs entirely in your browser.
Nothing you type leaves your browser. There is no upload, no server, and no storage — close the tab and it is gone.
Words
Reading time
Speaking time
characters left.
Over by characters.
What this word counter does
This is a live word and character counter. Type or paste text into the box and every figure updates immediately: words, characters with and without spaces, sentences, paragraphs, unique words, average word and sentence length, and an estimate of how long the text takes to read or to read aloud. Nothing is submitted anywhere — all the counting happens in JavaScript inside your own browser, so drafts, essays, and client work never leave your machine.
It is built for the moments when a length limit is the whole problem: a 500-word college essay, a 155-character meta description, a 2,000-word article commission, a conference talk that has to fit in ten minutes, or a personal statement with a hard character cap. Set the optional character limit and the tool tells you how much room is left — or how far over you have gone.
How to use it
- Paste or type your text. The counts start updating on the first keystroke; there is no button to press.
- Read the panel on the left for structural counts — characters, sentences, paragraphs, and the averages that hint at readability.
- Adjust the speeds on the right if the defaults do not match you or your audience. Reading and speaking rates vary a lot between people and between subject matter, so the estimate is only as good as the rate you give it.
- Set a character limit if you are writing to a cap, and watch the remaining count.
How each figure is calculated
- Words — any run of non-whitespace characters counts as one word. So well-known and don't each count as one, and a number like 1,250 counts as one. This matches how most word processors and submission systems count.
- Characters (with spaces) — the raw length of the text, including spaces, tabs, and line breaks. This is usually the number that social platforms and form fields enforce.
- Characters (no spaces) — the same total with every whitespace character removed. Some academic and translation briefs use this measure.
- Sentences — the text is split on
.,!,?, and the ellipsis character, and each remaining chunk that contains something other than whitespace counts as one sentence. - Paragraphs — every line that contains at least one non-whitespace character. Blank lines between paragraphs are ignored, so single-spaced and double-spaced drafts give the same answer.
- Unique words — words compared case-insensitively with surrounding punctuation stripped, so The, the, and the, count once between them. Comparing this to the total word count is a quick check for repetitiveness.
- Reading and speaking time — words divided by the rate you set, rounded to the nearest second: minutes = words ÷ words-per-minute. If you clear the rate or set it to zero, the tool shows a dash rather than an impossible number.
Sentence detection is deliberately simple, and that is worth knowing. Abbreviations that end in a period — "Dr. Chen", "e.g.", "approx." — are read as sentence endings, so a text full of them will report more sentences than it really has. The average-sentence-length figure inherits the same wobble. Treat both as guidance, not as a measurement.
A worked example
Suppose you paste in a blog draft and the tool reports 1,400 words, 8,120 characters with spaces, 78 sentences, and 24 paragraphs. At the default reading speed of 200 words per minute the estimate is 1400 ÷ 200 = 7 minutes exactly. Raise the rate to 250 and it drops to 5 min 36 sec; lower it to 150 — a reasonable pace for dense technical material — and it rises to 9 min 20 sec. That spread is the honest range, and it is why the rate is an input rather than a fixed constant.
The averages tell you something too: 1400 words across 78 sentences is roughly 18 words per sentence, which sits in the comfortable range for general web writing. If that number climbed past the high twenties you would be looking at a draft worth breaking up. And if unique words came back at 420 out of 1,400, roughly 30% of the words are distinct — normal for English prose, where function words like the, of, and to repeat constantly.
Choosing a reading speed
The 200 words-per-minute default is a deliberately round, conservative figure for silent reading of general prose. A 2019 meta-analysis of reading-rate studies by Marc Brysbaert put average silent reading of English non-fiction at roughly 238 words per minute, and reading aloud at roughly 183 — with wide variation between individuals and text types. Presenters typically speak more slowly than they read aloud in private, which is why the speaking default here is 150. Whatever your audience actually does, change the number: the arithmetic is trivial, and the rate is the part that carries all the uncertainty.