Word & Character Counter
Count words, characters, sentences, and paragraphs instantly. Estimate reading and speaking time. Free, browser-based, no server required.
- Runs in your browser
- Tool input is not uploaded to ZeroTool servers
- Free · No Sign-Up
Scan with WeChat to share this tool
Examples, details and FAQ Worked examples, how it compares with other tools, and answers to common questions.
Worked Examples
Each example shows the input and the numbers that the tool shows for it.
Abbreviations and the sentence count
Dr. Smith arrived at 3.30 p.m. He left early.
Words: 9
Sentences: 3
A human reader sees two sentences. The Unicode sentence rules also end one after Dr. because a capital letter follows it; the period inside 3.30 does not end a sentence. The three sentences are Dr., Smith arrived at 3.30 p.m. and He left early. Read the sentence count as an estimate when the text has abbreviations.
A short post with an emoji
I love it 👨👩👧
Characters: 18
Characters (no spaces): 15
Words: 4
The family emoji is three person emoji joined by two zero-width joiners: 8 UTF-16 code units. The emoji also counts as one word because whitespace separates it. X counts this emoji as 2 (see Limits), so X would count this post as 12.
English with a Japanese place name
Meeting at 東京 office
Words: 5
The two kanji in 東京 count as two words, and the three English words are split on spaces.
Paragraphs and line breaks
Release notes
Fixed the login bug.
Improved search speed.
Thanks to all testers!
Sentences: 4
Paragraphs: 3
The heading line counts as a sentence because a line break ends a sentence. The two lines in the middle are one paragraph: a single line break does not start a new paragraph, only a blank line does.
Character Counting Standards
Different platforms count characters differently. Some count Unicode code points, some count graphemes (what a reader sees as one character), and some weight characters. This tool counts JavaScript string length, which is UTF-16 code units. Most Latin, Cyrillic, Chinese, Japanese and Korean characters count as 1. Emoji, rare CJK characters outside the Basic Multilingual Plane (for example 𠮷) and letters written with a separate combining accent count as 2 or more.
How Each Number Is Counted
| Stat | Rule in this tool | Example |
|---|---|---|
| Words | Each Han, hiragana and katakana character is 1 word (half-width katakana included; a voiced sound mark belongs to its kana). All other text, Korean included, is split on spaces, tabs and line breaks | well-being, 3.14, and https://example.com/a are 1 word each; **bold** is 1 word with the asterisks; 今日は晴れです。 is 7 words |
| Characters | JavaScript string length (UTF-16 code units), including spaces and line breaks | 👍 counts as 2 |
| Characters (no spaces) | The same, with all whitespace removed | — |
| Sentences | Unicode sentence boundaries (UAX #29), computed by the browser's Intl.Segmenter. Sentence-ending marks include . ! ? and the full-width 。 ! ?; a period before a lowercase word or inside a number does not end a sentence, and each line break ends one | Dr. Smith agreed. counts 2; The value is 3.14. Next line. counts 2; 今日は晴れです。明日は雨です! counts 2 |
| Paragraphs | Blocks separated by a blank line; a line that has only spaces also counts as blank | — |
| Reading / speaking time | Per minute: 236 / 190 English and other words, 226 / 133 Korean words, 390 / 228 Chinese characters, 236 / 190 Japanese characters; parts are added and rounded up to whole minutes; under 1 minute shows < 1 min | 1,000 English words → 5 min reading, 6 min speaking; 390 Chinese characters → 1 min reading |
Other counters can give different numbers for the same text. Microsoft Word, Google Docs, and web counters do not all treat hyphenated words, URLs, Markdown symbols, and code in the same way. When you must stay under a limit, measure with the tool that the reviewer uses.
Limits
- Abbreviations and lowercase starts affect the sentence count.
Dr.before a capital letter counts as a sentence end. A sentence that starts with a lowercase letter or a digit after a period is not counted as a new sentence.…alone does not end a sentence. - Line breaks end sentences. A heading or list item on its own line counts as a sentence, and text with hard line breaks inside sentences (for example, copied from a PDF) shows more sentences than it has.
- Chinese and Japanese words are characters. The tool does not segment Chinese or Japanese into dictionary words, so
東京is 2 words, not 1. Punctuation is never a word, also when spaces separate it:a - bis 2 words and你好 , 世界 !is 4. - Reading times are estimates. All rates come from Brysbaert (2019), How many words do we read per minute?, Table 5, which gives averages over many studies: English 236 wpm silent and 190 aloud; Korean 226 and 133; Chinese 260 and 152 wpm. The paper often converted character counts to words at 1.5 characters per word; multiplying back gives about 390 and 228 characters per minute. Table 5 has no Japanese row, so a text that contains any kana uses 236 and 190 per character; this is a word-count estimate, not a measured Japanese reading speed.
- Characters are code units, not what you see. A thumbs-up with a skin tone (👍🏽) counts as 4, a family of four (👨👩👧👦) as 11, and an accent typed as a separate combining mark adds 1.
- Posts on X. X counts each URL as 23 characters and most CJK characters and emoji as 2 (counting characters), so the X count can differ from the Characters stat here.
1, 2, 3 — 4... ! ?
Words: 4
The dash, ! and ? stand alone between spaces and are not counted; 4... counts because it has a digit.
FAQ
What counts as a word?
Each Chinese character (Han), hiragana and katakana counts as one word, because Chinese and Japanese do not put spaces between words. Every other text, Korean included, is split on spaces, tabs and line breaks, and a piece made only of punctuation (such as a dash between spaces) is not a word: hyphenated terms like 'well-known', numbers like '42' and URLs count as one word. Mixed text adds the two: 'Meeting at 東京 office' is 5 words.
How is reading time calculated?
Each script has its own rate, and the parts are added. All rates come from Table 5 of a 2019 meta-analysis of reading rates (Brysbaert), silent reading for Reading Time and reading aloud for Speaking Time: English and other space-separated words 236 and 190 words per minute, Korean words 226 and 133, Chinese characters 390 and 228. Table 5 of that paper has no Japanese row, so in a text that contains any kana, kanji and kana use 236 and 190 per minute, counted like words. A nonzero time under one minute shows '< 1 min'; longer times round up to whole minutes, so 1,000 English words show 5 min reading and 6 min speaking.
What is the difference between characters with and without spaces?
Characters counts every UTF-16 code unit in the text, including spaces, tabs and line breaks. Characters (no spaces) removes every character that JavaScript treats as whitespace first, including the full-width space U+3000 and the no-break space. Invisible characters that are not whitespace, such as a zero-width space, are still counted.
Is there a length limit?
The tool sets no limit. Every input event counts the whole text again, so the practical limit is what your browser can handle in a text box.
Why does X say my post is too long when Characters shows 280 or fewer?
X does not count the way this tool does. X's documentation counts every URL as 23 characters and most CJK characters and emoji as 2. A family emoji such as 👨👩👧 counts as 2 on X but as 8 here, because this tool counts UTF-16 code units. Check the final post in the X composer.
Is my text sent to a server?
No. The counting runs in your browser tab with JavaScript; the text is not sent, and the tool does not save it in browser storage. Page analytics record one 'count' event with the tool name when you leave the text box after an edit, without the text.