Word Counter

How to use the word counter

Paste or type into the box and every figure updates as you go — words, characters, characters without spaces, sentences, paragraphs, reading time, and speaking time. If you are writing to an assignment length, the big number at the top is the one to watch. A word here is simply a run of characters with whitespace on either side, which is the rule word processors use and the only definition that behaves predictably: “well-known” counts once, “it’s” counts once, and “New York” counts twice. Because the rule is purely about whitespace, a figure written 1,250 is one word, a date written 12 March 2026 is three, and a lone em dash floating between two spaces is a run of its own and counts as one. That last case is the price of a definition with no exceptions, and it is exactly why the count never drifts from what a word processor would tell you.

Essay portals and application forms are often capped in characters rather than words, so the panel reports both — every character including spaces and line breaks, and the same text again with all whitespace stripped out. Characters are counted by code point rather than by the storage unit a browser uses internally, which is why an emoji, an accented letter, or a Japanese character counts as one here. A naive counter measures the UTF-16 length instead and reports a single emoji as two, telling you that you have blown a 280-character limit when you have not. Sentences close at a period, question mark, exclamation point, or ellipsis, and a run of marks closes one sentence rather than several, so “Wait!!” counts as one. Paragraphs split on blank lines only; a single line break inside a paragraph is treated as a soft wrap and will not inflate the count.

The reading estimate runs at 238 words per minute by default, the average adult silent reading rate for English non-fiction reported in Marc Brysbaert’s 2019 review in the Journal of Memory and Language — a meta-analysis of 190 studies covering more than 18,000 readers. The same review put fiction nearer 260 words per minute, because its words are shorter, and reading aloud at about 183. The speaking figure this tool uses is lower still, at 130 words per minute: that is not a research number but the pace presentation coaches quote as comfortable, deliberately slower than a read-aloud test because a talk contains pauses, breaths, and questions that a lab task does not. Five minutes of stage time therefore comes to roughly 650 words, which is a useful sanity check before you write the script rather than after. Both estimates always round up, so a two-line note reads as 1 min rather than 0 min.

Nothing you paste leaves your browser. The counting happens in the page itself — no upload, no server round trip, nothing stored between visits, and the text is never written into the page URL, so there is no shareable link quietly carrying your draft around. That matters when the text is an unpublished manuscript, a cover letter, a legal filing, or a client document you are not free to share. The reading speed is an editable field too, so if you read faster or slower than the 238 words per minute default, or you are estimating for a particular audience — children, a specialist readership, someone reading in a second language — type the number you want and the reading time follows. Dense technical prose pulls real reading rates well below the average, and the honest move when a piece is difficult is to lower the figure rather than quote the default and hope.

The formula

The counts themselves are a straight tally of what you typed — words are whitespace-separated runs, characters are code points, sentences close at terminal punctuation, and paragraphs are blocks separated by a blank line. None of that needs arithmetic, which is why the counter can keep up with you on every keystroke.

Only the two time estimates are calculated, and both are the same division followed by a ceiling. The ceiling is why the tool never reports zero minutes for text that has any words in it at all, and it is the reason a 100-word note and a 200-word note can both read as 1 min while their speaking times differ.

reading time = words ÷ reading speed, rounded up
speaking time = words ÷ 130 wpm, rounded up
characters = count of code points, spaces and line breaks included
characters without spaces = the same count with all whitespace removed

Worked example with the sample text the tool opens on. It is 53 words, 303 characters, 250 characters without spaces, 3 sentences, and 2 paragraphs. Reading time is 53 ÷ 238 = 0.22 minutes, which rounds up to 1 min; speaking time is 53 ÷ 130 = 0.41 minutes, which also rounds up to 1 min. Take a longer piece and the two diverge sharply: a 1,500-word article is 1,500 ÷ 238 = 6.30 minutes of silent reading, so 7 min, but 1,500 ÷ 130 = 11.54 minutes read aloud, so 12 min. The same words take roughly 1.8 times as long to say as to read.

The 238 comes from Marc Brysbaert’s 2019 paper in the Journal of Memory and Language, volume 109, which reviewed 190 studies and more than 18,000 participants and put average adult silent reading of English non-fiction at 238 words per minute. Fiction ran nearer 260 in the same review, because fiction uses shorter and more familiar words, and oral reading — reading a passage aloud as fast as is comfortable — came out at about 183 words per minute across 77 studies. Those are means across large samples, not a promise about any individual reader, and the spread between people is wide. That is precisely why the reading speed here is an editable field rather than a fixed constant.

The 130 words per minute used for speaking time is a different kind of number, and it is worth being clear that it is a convention rather than a research finding. It sits deliberately below Brysbaert’s 183 wpm oral reading figure because a presentation is not a read-aloud test: a speaker pauses, breathes, waits for a laugh, repeats a point, and answers a question. Budgeting a talk at 130 words per minute puts a 5-minute slot at about 650 words and a 20-minute slot at about 2,600, and coming in slightly short is almost always the safer error. If you have timed yourself and you genuinely deliver faster, treat the tool figure as a ceiling and write to your own measured pace.

Frequently asked questions