Word Counter: Words, Characters, Reading Time and Keywords

Words are runs of text separated by whitespace or by an em or en dash. Punctuation around a word is ignored, apostrophes and hyphens inside a word stay ("don't" and "well-known" are one word each), and a token with no letter or digit, such as "-" or "&", is not counted. Chinese, Japanese, Thai and other scripts written without spaces are split with the browser's word segmenter when it has one, and character by character when it does not.

Characters are what a reader sees as one character, not what JavaScript reports as .length. An emoji with a skin tone, a flag or an accented letter written as two code points each counts once, and a Windows line break (CRLF) counts as one. The tool also shows the count without spaces and the size in UTF-8 bytes, which matters for databases and APIs that limit bytes instead of characters.

Sentences end at a full stop, question mark, exclamation mark or ellipsis followed by a space, a closing quote or the end of the text. Decimals, domain names and common abbreviations such as Dr. and e.g. do not end a sentence, but no rule set can be right for every abbreviation in every language, so treat the sentence count as a good estimate. Paragraphs are blocks separated by blank lines.

Reading time uses 238 words per minute, the average silent reading rate of adults reading English non-fiction in the 2019 meta-analysis by Marc Brysbaert (190 studies, 18,573 participants). Speaking time uses 150 words per minute, a common planning pace for talks and voice-overs; that figure is a convention, not a measured average. Both are estimates: technical text, a second language or a slide deck read aloud with pauses will run slower.

You can scale the numbers yourself. A 1,000-word article is about four minutes of reading and about six and a half minutes of speaking at these rates.

The keyword list shows the most frequent words or two- and three-word phrases, with each one's share of all words. English stop words such as "the" and "of" can be skipped. Use it to spot repetition in your own writing. The tool does not claim that any percentage helps a page rank, and it does not suggest a target: write for the reader first.

Character limit presets show how much room is left. X posts are weighted the way X documents it (most Latin characters count 1, Chinese, Japanese, Korean and emoji count 2, every link counts 23), which this tool approximates. Google shortens titles and descriptions by pixel width, so 60 and 155 characters are planning guides rather than rules. SMS messages switch from GSM-7 (160 characters, or 153 per part) to Unicode (70, or 67 per part) as soon as one character falls outside the GSM set, and the tool shows which applies.

Pasting a text of several megabytes works: counting starts shortly after you stop typing and runs in chunks so the page stays responsive. Select part of the text to see its own counts.