"How long should this be?" gets answered in at least three completely different units depending on where the text is going: a college essay wants a word count, a tweet wants a character count, and a meta description wants something closer to a pixel width. Mixing these up is an easy way to submit an essay that's technically under the word limit but reads short, or a tweet that gets cut off mid-sentence. Here's how each measurement actually works.
Word count: simpler than it sounds, until it isn't
At its core, a word counter splits text on whitespace and counts the pieces. That's straightforward for most writing, but a few edge cases trip people up:
- Contractions — "don't," "it's," and "can't" each count as one word, not two, since there's no space inside them.
- Hyphenated words — "well-known" and "state-of-the-art" also count as single words for the same reason.
- Numbers and standalone punctuation — "3.14" typically counts as one word. A lone em dash or ellipsis with spaces on both sides is where tools disagree: a simple whitespace-based counter (including this site's) counts it as a word, while some word processors skip standalone punctuation.
- Em dashes without spaces — "wait—really?" written with no spaces around the dash can accidentally merge into a single counted word, quietly lowering your count if you use this style often.
Most academic and publishing word limits (essays, articles, submissions) are genuinely asking for this whitespace-split word count, so it's the right measurement to check against for that use case.
Character count: where it gets platform-specific
Character limits are stricter and less forgiving than word limits, but "character" doesn't mean the same thing everywhere:
- Simple character count counts every letter, number, space, and punctuation mark as one character each — this is what most meta description and SMS-length guidance refers to.
- X (Twitter) uses a weighted count — most Latin characters count as 1, but CJK characters and emoji count as 2, and any URL counts as a fixed 23 characters regardless of how long the actual link is, according to X's developer documentation. A post that looks short can still hit the limit fast if it's full of links or non-Latin text.
- "Characters without spaces" is a separate figure some tools report alongside the standard count, mainly used for print-length estimates (like typesetting) rather than platform limits.
The practical takeaway: check which kind of character count a platform actually enforces before trusting a generic counter's number against it — a simple count and X's weighted count can diverge meaningfully on link- or emoji-heavy text.
Reading time: an estimate, not a measurement
Unlike word and character count, reading time isn't something a tool measures directly — it's calculated by dividing your word count by an assumed reading speed, commonly 200–250 words per minute for adult silent reading (this site's Word Counter defaults to 238, with 200 as a slower option), then rounding to a whole number of minutes. That means two tools can report different reading times for identical text if they assume different speeds; neither is "wrong," they're just using different baseline assumptions. Treat reading time as a rough guide for pacing a piece, not a precise figure.
Which measurement should you actually check?
- Essays, articles, cover letters, submissions → word count, almost always the simple whitespace-split version.
- Tweets/X posts, SMS, meta titles and descriptions → character count, and check whether the platform weights certain characters differently.
- Blog posts, newsletters, long-form content → word count for the limit itself, reading time as a secondary signal for how the piece will feel to a reader.
Related text tools worth knowing
If you're prepping text for a limit, a few other tools on this site pair naturally with a word/character check: Case Converter for fixing capitalization before a final count, Text Diff for seeing exactly what changed between two drafts, and Find & Replace for bulk-fixing a phrase across a long document before you count it.
Words, characters (with and without spaces), sentences, paragraphs, and reading time — calculated instantly as you type.
Open Word Counter →What counts as one "character"?
To a person, one visible symbol is one character. To software, it depends on what's being counted. There are three common measures, and they can give three different numbers for the same text:
- Grapheme clusters: what a reader sees as one character. This is what people usually mean.
- Unicode code points: the individual encoded characters. Some visible symbols are built from several of them.
- UTF-16 code units: what JavaScript's
.lengthcounts. Characters outside the Basic Multilingual Plane, including most emoji, take two units.
| Text | What you see | Code points | JavaScript .length | UTF-8 bytes |
|---|---|---|---|---|
| a | 1 | 1 | 1 | 1 |
| é (precomposed) | 1 | 1 | 1 | 2 |
| € | 1 | 1 | 1 | 3 |
| 😀 | 1 | 1 | 2 | 4 |
| 👨👩👧 (family emoji) | 1 | 5 | 8 | 18 |
The family emoji is three person emoji joined by invisible zero-width joiners. That's why it counts as 5 code points. Modern browsers can count true visible characters with Intl.Segmenter, which splits text into graphemes as defined by the Unicode standard.
SMS limits and segments
Text messages have their own rules. A single SMS holds up to 160 characters in the standard GSM 7-bit alphabet. If your message includes any character outside that alphabet, such as an emoji or many accented letters, the whole message switches to UCS-2 encoding and the limit drops to 70 characters. Longer messages are split into segments of 153 (GSM) or 67 (UCS-2) characters, because each segment needs room for a header that joins them back together. Some GSM symbols, like €, [, ], {, }, ^, ~, | and \, use an escape code and count as two characters.
Bytes: when storage is the limit
Databases, APIs, and some form fields limit length in bytes, not characters. In UTF-8, basic Latin letters take 1 byte, most accented Latin letters take 2, many symbols like € take 3, and emoji take 4 or more. A 255-byte field can hold 255 English letters but only 63 four-byte emoji.
Languages without spaces
Word counts based on spaces don't work for Chinese, Japanese, or Thai, which don't put spaces between words. A space-splitting counter will report a whole sentence, or even a paragraph, as one "word." For these languages, character count is the usual length measure. Tools that need real word boundaries use dictionary-based segmentation, like Intl.Segmenter with granularity: "word".
Titles and descriptions in search results
Search engines don't publish a strict character limit for page titles or meta descriptions. Google's documentation says snippets are truncated as needed to fit the device width. Because wide letters like "W" take more space than narrow ones like "i", a pixel budget can fit different numbers of characters. Character counts in the 50–60 range for titles and about 150–160 for descriptions are popular rules of thumb, not official limits. Put the important words first so a cut-off doesn't hide them.
Cleaning text before you count
Invisible characters change counts too. Non-breaking spaces, doubled spaces, and trailing line breaks pasted from a PDF or word processor can inflate a character total. Running the text through the Text Cleaner first gives a count that matches what readers will actually see.
Frequently asked questions
Does a word count tool count contractions as one word or two?
Almost all word counters, including this one, count a contraction like "don't" as a single word, since it's split by whitespace rather than by the apostrophe. The same applies to hyphenated words like "well-known," which are also typically counted as one word.
Why does Twitter/X count some characters as more than one?
X uses a weighted character count where most Latin characters count as 1, but CJK (Chinese, Japanese, Korean) characters and certain symbols count as 2, and URLs are automatically shortened to a fixed length (currently 23 characters) regardless of the actual link length.
How is reading time calculated from a word count?
Reading time is typically estimated by dividing the total word count by an average adult silent reading speed, commonly assumed to be around 200–250 words per minute, then rounding to a whole number of minutes.
Why does one emoji count as two characters in some tools?
Many tools, including JavaScript's .length, count UTF-16 code units. Most emoji sit outside the Basic Multilingual Plane and need two code units, so they count as 2 even though they appear as one symbol. Combined emoji, like family or flag sequences, can count even higher.
How many characters fit in one SMS?
Up to 160 characters using the standard GSM 7-bit alphabet, or 70 if the message contains an emoji or other character outside that alphabet. Longer messages are split into segments of 153 or 67 characters each.