What's the difference between characters and words?
A word is a group of letters or digits between spaces; a character is every single letter, digit, symbol and space. Which one matters depends on who sets the limit.
Published November 4, 2026
Words and characters at a glance
Take this short sentence:
The café opens at 9 a.m. 👍
Counted the way Microsoft Word does it, that is 6 words: anything between spaces that contains at least one letter or digit counts as one word, so “9” and “a.m.” count, and the emoji on its own doesn’t. The same sentence has 26 characters with spaces and 20 characters without spaces. Behind the scenes it is also 26 Unicode code points, 27 UTF-16 code units and 30 UTF-8 bytes.
All of those numbers are correct. They just measure different things, and different systems use different ones.
Characters with and without spaces
Characters with spaces counts every letter, digit, punctuation mark, symbol and space you can see. CountWords, like most counters, leaves line breaks out of this number. Characters without spaces drops all whitespace: spaces, tabs and special spaces such as no-break spaces.
When a limit says only “characters”, assume spaces count unless the instructions say otherwise. Online forms count everything you type, including spaces, and so do most application portals and social networks. Characters without spaces turn up in some publishing, translation and academic guidelines, where they give a measure of text that doesn’t depend on spacing.
Code points, UTF-16 units and bytes
What looks like one character on screen can be stored as several pieces. Software often counts those pieces rather than what you see, which is why a form can reject text that looks short enough.
| Text | Characters you see | Code points | UTF-16 units | UTF-8 bytes |
|---|---|---|---|---|
| A | 1 | 1 | 1 | 1 |
| é (single character) | 1 | 1 | 1 | 2 |
| é (e + combining accent) | 1 | 2 | 2 | 3 |
| 中 | 1 | 1 | 1 | 3 |
| 👍 | 1 | 1 | 2 | 4 |
| 👍🏽 (with skin tone) | 1 | 2 | 4 | 8 |
| 👨👩👧 (family) | 1 | 5 | 8 | 18 |
| 🇯🇵 (flag) | 1 | 2 | 4 | 8 |
- Code points are Unicode’s building blocks. Accents can be separate code points, and many emoji are sequences of several code points joined together.
- UTF-16 code units are how JavaScript, Java and Windows store text. Most emoji take two units each.
- UTF-8 bytes are how most web pages, files and databases store text: 1 byte for basic Latin letters, 2 for most accented Latin, Greek, Cyrillic, Arabic and Hebrew letters, 3 for Chinese, Japanese, Korean and Thai, and 4 for most emoji.
Two strings that look identical can even give different counts: “é” typed on one keyboard may be a single code point, while text copied from elsewhere may use “e” plus a combining accent.
When each count matters
Essays, articles and applications
Most writing limits are in words: essays, articles, abstracts and cover letters. Some application systems use characters instead, usually with spaces. If the instructions are unclear, ask, or check the counter of the system you will submit through.
Web forms
The HTML standard measures a text field’s maxlength in UTF-16 code units, so an emoji can use two of your allowed characters and a family emoji eight. Browsers haven’t always agreed on this, and server-side checks may count differently again, so test with the text you expect people to enter.
Databases
- MySQL counts
VARCHAR(n)in characters (code points), but the oldutf8character set can’t store emoji or other 4-byte characters; useutf8mb4. - PostgreSQL counts
varchar(n)in characters. - Oracle
VARCHAR2columns can be sized in bytes or in characters, and byte semantics is the default, so a column that holds 100 English letters may hold far fewer Chinese characters. - SQL Server sizes
nvarchar(n)in byte-pairs, that is, UTF-16 code units.
Social media and SMS
Platforms each pick their own rule. X counts a weighted length in which emoji and Chinese, Japanese and Korean characters count as 2 and each link as 23. Threads counts each emoji by its UTF-8 size. TikTok’s caption limit is defined in UTF-16 units in its posting API. For Instagram and LinkedIn, CountWords counts characters as you see them. SMS uses yet another scheme, based on the GSM-7 alphabet; see the SMS character limit guide.
Translation
Translators price and plan work in different units. Word counts of the source text are common for English and other languages written with spaces. For Chinese and Japanese, which don’t use spaces between words, character counts are the norm. In Germany, the Normseite (standard page) of 1,800 characters including spaces is a common unit. Translated text also changes length, so a translation can overflow a field that was sized for the original language.
Converting words to characters
There is no fixed ratio: it depends on the language, the subject and your writing style. Technical writing with long terms has more characters per word than a children’s story. The reliable way is to measure your own text: divide the characters with spaces by the number of words. If your text averages 6 characters per word including the space, a 4,000-character limit is roughly 670 words.
See every count at once
Paste your text into the counter below. The character view shows characters with and without spaces, words, Unicode code points, UTF-16 code units and UTF-8 bytes side by side, so you can match whichever count your form, database or platform uses. Everything is counted in your browser and nothing is uploaded.