How Many Words Is 4,000 Characters? Characters, Words and Limits

A form tells you that you have 4,000 characters. You paste in what you wrote, and it says you are 300 over. You cut a paragraph. Now a different tool, checking the same text, says you were never over at all.

Both tools are working correctly. They are counting different things, and neither of them told you which.

Characters to words, at a glance

For ordinary English prose, 4,000 characters is about 650 words.

That ratio is not a rule of thumb borrowed from somewhere else. It is measured: across the 12,690 words published on this site, the average word runs 5.14 letters, which becomes 6.14 characters once you add the single space that follows it. The commonly quoted average word length in English is five letters, and the gap is simply that ordinary prose contains longer words than a dictionary average suggests.

Going the other way: 500 words is roughly 3,070 characters, and 1,000 words about 6,140.

One row in that table has enough behind it to need a page of its own. 3,000 characters is the LinkedIn post limit, which is why it comes up more than the others — how many words is 3,000 characters covers the conversion, what actually fits in the space, and how to cut a post down to it.

How many words is 4,000 characters?

About 650 words, or roughly a page and a half of typed prose.

This is the limit people meet most often, because it is the size of the UCAS personal statement used for UK university applications. If you are writing to that limit, two things are worth knowing before you start cutting. The 4,000 includes spaces, and it includes line breaks — so adding a blank line between paragraphs costs you characters. And most application forms count what you paste, not what you typed, so curly quotes carried over from a word processor arrive intact and count the same as straight ones.

How many words is 500 characters?

About 81 words — a single dense paragraph, or the opening of a cover letter.

How many words is 1,000 characters?

About 163 words, which is three or four ordinary paragraphs.

How many words is 2,000 characters?

About 326 words, roughly a printed page of a book.

How many characters is 500 words?

About 3,070 characters including spaces. For 1,000 words, about 6,140.

The ratio holds for normal prose and breaks in predictable places. Technical writing with long nouns runs higher — "authentication" alone is 14 characters. German runs much higher. Text with a lot of short function words runs lower. If you need the real number rather than an estimate, paste it into a character counter and read it off.

Do spaces count as characters?

Yes — in essentially every field that enforces a limit. SMS, X, Instagram bios, meta descriptions, database columns, application forms. If a computer is measuring it, a space is a character like any other, because it takes up a byte like any other.

This is the single most common reason a count surprises you. The sentence "Do spaces count as characters" is 29 characters with spaces and 25 without, and both numbers are correct — they answer different questions.

The rule in practice, which almost nothing states out loud:

• A limit includes spaces unless it explicitly says otherwise. Assume it does.
• Counts without spaces are for humans, and mostly for academic work — some universities set essay limits that way so that spacing cannot be gamed.

So if you are checking text against a limit, the number you want is nearly always the one with spaces. The count without spaces is for when a style guide or a supervisor has specifically asked for it.

One detail if you write in a plain-text editor: most tools reporting a count "without spaces" actually strip all whitespace, so tabs and line breaks vanish from that total too. The counter on this site does exactly that. A document made of many short lines will show a bigger gap between its two numbers than the spaces alone would explain.

Character limits on every platform

The pattern across all of these is worth stating before the numbers: the stated limit and the visible limit are rarely the same. Most platforms accept far more than they display.

TikTok caption

This one deserves an honest answer rather than a confident one. TikTok does not publish a caption limit in its own help documentation, and the sites that do publish a figure disagree with each other — 2,200 characters is the most commonly repeated number, but 4,000 appears just as often, and the limit has been raised more than once.

So the only reliable answer is the one you can check in thirty seconds: open the caption field, paste a block of text longer than you need, and see where it stops accepting characters. That is the number for your app version and region today, which is more than any article can promise.

What is stable is the part that matters for reach. Only the first line or two appear over the video before the caption is truncated, so the opening is doing nearly all the work regardless of how much the field will hold.

Instagram bio

150 characters, and line breaks count against it. This is the tightest limit of the group, which is why bios read like telegrams.

YouTube title and description

Descriptions allow a maximum of 5,000 characters, which is stated in YouTube's own documentation. Roughly the first 150 appear above the fold before the "more" button hides the rest.

Titles are widely reported at 100 characters, though YouTube does not state the figure in the same place. What is worth planning around either way is that search results and suggested-video panels show only about 60, so the back half of a long title is invisible where it counts.

LinkedIn headline

220 characters, truncated far earlier almost everywhere it is displayed — in search results and comment threads you can expect well under half of it to show.

Meta description

Nearly every SEO article gives a number, usually 155 or 160. Google's own documentation gives none.

The Search Central guidance on snippets states there is no limit on how long a meta description can be, and that the snippet is truncated in search results as needed, typically to fit the width of the device.

That last phrase matters. Truncation happens by rendered width, not by character count, so it varies from phone to laptop, and a description full of wide characters is cut sooner than one full of narrow ones. The familiar 155-to-160 figure is an observed average of where desktop results tend to cut. It is not a specification. Write to about 160 because that is where truncation usually lands, but put what matters in the first 100 characters and let the tail be the part that can safely disappear.

Why does my SMS say 70 characters?

Because one character in your message is not in the SMS alphabet, and that changes the encoding for the entire message.

A text holds 160 characters using GSM-7, the compact 7-bit character set SMS was designed around. GSM-7 is small. It has room for the straight apostrophe and not the typographic curly one. It has no em dash, no curly quotes, no emoji.

The moment a message contains one character outside that set, the whole message switches to UCS-2 encoding and the limit drops from 160 to 70 — for the entire message, not just the offending character.

This is why a message that looked fine in your notes app arrives as three texts. Autocorrect on both iOS and Android converts a straight apostrophe into a curly one as you type. One invisible substitution inside the word "don't", and your 160 characters became 70.

Longer messages are split into segments, and splitting costs extra: concatenated messages carry a header for reassembly, leaving 153 characters per segment in GSM-7 and 67 in UCS-2. So a 200-character plain message is two segments. Add a single emoji and the same message becomes four.

If you are sending anything where the segment count matters — an appointment reminder, anything billed per message — write it somewhere with smart quotes turned off, and check it before it goes.

Why does an emoji count as 2 characters?

Add one emoji and the counter jumps by two. This is not a bug or a rounding error, and it happens for two genuinely different reasons that arrive at the same answer.

In the browser, it is about how text is stored. JavaScript measures a string in UTF-16 code units. Characters up to U+FFFF take one unit each. Emoji sit far above that ceiling, so each is stored as a surrogate pair — two units — and any counter built on the standard string length, including the one on this site, reports 2. This is not limited to emoji: many CJK extension characters and historic scripts behave identically.

On X, it is a deliberate policy. The published weighting configuration sets a maximum weighted length of 280 and a default character weight of 200, on a scale where 100 means one character. Four Unicode ranges are granted weight 100, the main one covering code points 0 to 4351 — Latin, Greek, Cyrillic, Hebrew, Arabic and most European accented letters.

Everything outside those ranges — every emoji, and all Chinese, Japanese and Korean text — carries the default weight of 200, and therefore counts as two. A 280-character post is 280 Latin characters or 140 Japanese ones.

Three consequences worth carrying around:

• Accented Latin is free. European languages get the full 280, because those letters sit inside the weight-100 range.
• A family emoji costs two, not fourteen. Sequences like the four-person family are several people joined by invisible zero-width joiners, seven code points in total. X parses the whole sequence as one emoji and charges 2 for it. A browser counter using raw string length charges 11. This is the practical difference between counting code points and counting what typographers call grapheme clusters — the units a reader would call "characters".
• Which means that, for once, X is more generous than the counter in your browser.

Does a link count as characters on Twitter?

Yes, but at a fixed price. The same configuration file sets the transformed URL length at 23.

Every URL you post is replaced by a wrapped link of fixed size, so a 200-character tracking URL stuffed with campaign parameters costs exactly what a bare domain costs: 23 characters, plus the space in front of it.

There is no benefit to shortening a link before posting it to X. Shorteners are worth using for click tracking, and for nothing else here.

When your character count does not match

If a form and a counter disagree about the same text, it is almost always one of five things:

• Spaces or line breaks are being counted by one and stripped by the other.
• An emoji or non-Latin character is costing two where you expected one.
• A URL is being recounted at a fixed length rather than its real one.
• A curly apostrophe has changed the encoding underneath you, on SMS.
• The form counts what you paste, including the trailing newline that came with it. A stray blank line at the end of a pasted block is a real character and a common cause of being "one over".

Text boxes in application forms are the worst offenders, because many validate on the raw pasted value while showing you a count computed after trimming.

A note on password fields

Character counting turns up in one more place where the rules are different.

Length is what makes a password hard to guess, which is why current guidance favours long passphrases over short strings of symbols. But a password field is one of the few places a character counter should not be trusted to tell you whether your input is acceptable. Some systems silently truncate a long password rather than rejecting it. Some count bytes rather than characters, so an accented letter or an emoji quietly consumes more of the budget than it appears to.

The practical answer is not to count at all, but to let software generate and store it. That is covered separately in our guide to password managers.

Checking your own text

For a quick answer, paste into the character counter for characters with and without spaces, or the word counter when the limit is expressed in words.

Both run entirely in your browser. Nothing you paste is uploaded anywhere, which matters more than it sounds for the two things people most often paste into a counter: a personal statement, and a message they have not sent yet.