Word Count, Explained
Word count explained: how words and characters are counted, words per page, reading and speaking time, and the limits for posts, meta descriptions and essays.
By Udit Agarwal · Updated October 17, 2026
A word count is the number of words in a text, and most counters treat a word as any run of characters between spaces or line breaks. A character count is the number of letters, digits, symbols and spaces, usually given twice: with spaces and without. Together they decide whether a post fits on X, whether an essay is inside its limit, and roughly how many pages or minutes a text will take.
The usual rule is simple: split the text wherever there is a space, tab or line break, and count the pieces. POSIX, the standard behind the Unix wc command, defines a word exactly that way, as a non-empty string of characters delimited by white space. Punctuation stays attached to the word beside it, so "dog." is one word.
That rule explains the cases that surprise people:
| Text | Words | Why |
|---|---|---|
well-known |
1 | There is no space inside a hyphenated word |
don't |
1 | A contraction has no space |
3.14 or 1,000 |
1 | A number is one run of characters |
https://example.com/a-b-c |
1 | A web address has no spaces |
word – word |
3 | A dash with spaces around it counts as a word; closed up as word–word, it is one |
Programs disagree about these edge cases, so totals can differ slightly between a word processor, an online counter and a submission portal.
Languages written without spaces between words, such as Chinese, Japanese and Thai, break the whitespace rule completely: a whole sentence can come out as one word. Programs use their own rules for these scripts, and limits for Chinese and Japanese text are often set in characters instead.
A count with spaces includes everything: letters, digits, punctuation, spaces, tabs and line breaks. A count without spaces removes all of the whitespace. "Hello world." is 12 characters with spaces and 11 without. The pangram "The quick brown fox jumps over the lazy dog." is 44 characters, and its 8 spaces bring it to 36 without them. When a limit just says "characters", it nearly always means characters with spaces.
One visible character can be several stored code points: the British flag is two regional-indicator symbols, and a family emoji is several people joined by invisible joiners. A careful counter counts what the reader sees, one character each, but some systems count the hidden pieces, so the same text can measure differently on different platforms.
Size in bytes is a third measure. In UTF-8, the standard text encoding on the web, a plain English letter takes 1 byte, an accented letter such as é usually takes 2, most Chinese and Japanese characters take 3 and a typical emoji takes 4. "日本語" is 3 characters but 9 bytes. Databases and APIs often limit bytes rather than characters, so a 255-byte field holds 255 English letters but only 63 four-byte emoji.
A page of 12-point text with 1-inch margins holds about 500 words single-spaced and about 250 double-spaced. Treat those as midpoints. Depending on the font, size, margins and how many headings and short lines the text has, a single-spaced page holds roughly 400 to 600 words and a double-spaced page roughly 200 to 300. A wide font such as Arial fits fewer words than Times New Roman, and dialogue or bullet points leave more white space.
| Words | Single-spaced | Double-spaced |
|---|---|---|
| 250 | about half a page | about 1 page |
| 500 | about 1 page | about 2 pages |
| 1,000 | about 2 pages | about 4 pages |
| 1,500 | about 3 pages | about 6 pages |
| 2,000 | about 4 pages | about 8 pages |
| 3,000 | about 6 pages | about 12 pages |
| 5,000 | about 10 pages | about 20 pages |
Across the full range, 1,000 words fills anywhere from about 1.7 to 2.5 single-spaced pages, or 3.3 to 5 double-spaced.
Reading time is the word count divided by a reading speed in words per minute (wpm). A 2019 review of 190 reading studies by the psychologist Marc Brysbaert put the average adult reading English silently at about 238 wpm for non-fiction and 260 wpm for fiction, and most estimates for adults fall between about 200 and 300 wpm. Technical text and unfamiliar subjects are read more slowly. Typing tests use a different definition again, a "word" of exactly five characters, as the guide to typing speed explains.
Speaking is slower than reading. People presenting typically talk at about 120 to 160 wpm, and 130 wpm is a common planning figure for a speech. A five-minute talk therefore needs roughly 600 to 800 words, and a ten-minute talk roughly 1,200 to 1,600.
| Words | Reading at 200 wpm | Reading at 250 wpm | Speaking at 130 wpm |
|---|---|---|---|
| 500 | 2.5 min | 2 min | 3.8 min |
| 1,000 | 5 min | 4 min | 7.7 min |
| 1,500 | 7.5 min | 6 min | 11.5 min |
| 2,000 | 10 min | 8 min | 15.4 min |
| 3,000 | 15 min | 12 min | 23.1 min |
Most limits are set in characters, and a few in words. These are the ones people check most often:
| Where | Limit |
|---|---|
| Post on X (standard account) | 280 characters. Every link counts as 23, and emoji and Chinese, Japanese or Korean characters count as 2 |
| Text message (SMS) | 160 characters, or 70 if the message contains an emoji or any character outside the GSM 7-bit alphabet. Longer messages are split into parts |
| Meta description | No official limit, but Google shortens snippets to fit the screen, so about 150 to 160 characters is the usual target |
| Page title | Google shortens long titles in results; about 50 to 60 characters usually shows in full |
| LinkedIn post | 3,000 characters |
| Instagram caption | 2,200 characters |
| Common App personal essay | 250 to 650 words |
| UCAS personal statement | 4,000 characters, spaces included |
For meta descriptions, Google's snippet documentation says there is no length limit; the cut-off happens on display. For essays and assignments, read the rules on what counts: titles, headings, footnotes, captions, quotations and reference lists are included by some institutions and left out by others.
AI language models split text into tokens, which are common chunks of text: a short word is often one token, while a long or rare word is split into several. Each model family has its own tokenizer, so the same paragraph comes to a different number of tokens in different models, and context windows and API prices are measured in tokens.
A common rule of thumb for English is about four characters per token, and that is the estimate the ArrayKit counter shows: characters with spaces divided by 4, rounded up. "Hello world." is 12 characters, so about 3 tokens, and a 2,000-character text is about 500. The rule fits ordinary English prose best. Chinese, Japanese and Korean text, long numbers and unusual words usually take more tokens than it suggests, so when you are close to a limit, use the Token Counter, which runs OpenAI's tokenizers for exact counts and labels its Claude and Gemini figures as estimates.
The Word & Character Counter updates as you type or paste and shows nine figures: characters with spaces, characters without spaces, words, sentences, paragraphs, lines, size in UTF-8 bytes, an approximate token count and a reading time. This is exactly how it counts:
., !, ? or …, or the full-width marks used in Chinese and Japanese. Text after the last mark counts as one more sentence, and a line break does not end one. Abbreviations and decimals add to the total: "Dr. Smith arrived at 3 p.m. today." counts as four sentences.Below the figures, a Top words list shows the eight most frequent words, ignoring capital letters, one-letter words and common words such as "the", "and" and "of". It does not merge forms of a word, so "cat" and "cats" are counted separately; the Character Frequency Counter gives a full table of every word or letter. Copy stats copies all the figures and the top words as plain text.
The reading speed is fixed at 200 wpm, at the slow end of the research averages, so the estimate leans long. There is no speaking-time figure, so divide the word count by about 130 for that. For a PDF, the PDF Word Counter reads the file page by page, counts each Chinese or Japanese character as a word and adds a speaking time at 130 words a minute. For syllables and reading grade levels, use the Readability Score Checker. The counting happens in your browser, and the text you paste is not uploaded.
How many words is a page? About 500 words single-spaced and 250 double-spaced, assuming 12-point text and 1-inch margins. Depending on the font and layout, the range is roughly 400 to 600 words single-spaced and 200 to 300 double-spaced.
How long does it take to read 1,000 words? About four to five minutes for an adult reading silently at 200 to 250 words per minute. Reading the same text aloud at 130 words per minute takes nearly eight minutes.
How many words is a 5-minute speech? Roughly 600 to 800 words, at a speaking pace of 120 to 160 words per minute. At a steady 130 words per minute, plan on about 650.
Do spaces count as characters? In most limits, yes. "Characters" on its own normally means characters with spaces, and X, text messages and the UCAS personal statement all count them. A counter that shows both figures lets you check either kind of limit.
Is a hyphenated word one word or two? A counter that splits text at spaces, like the ArrayKit one, counts "well-known" as one word because there is no space in it. Some programs split at hyphens, so totals can differ slightly from tool to tool.
Why do two word counters give different totals? They treat dashes, hyphens, numbers, web addresses and text without spaces differently, and some skip headers, footnotes or captions. The differences are usually small; when a limit matters, trust the count of the system that enforces it.
How many characters can a post on X have? 280 for a standard account. Every link counts as 23 characters however long it is, and emoji and Chinese, Japanese or Korean characters count as two, so a post can reach the limit sooner than a plain character count suggests.