Count words, characters, sentences and paragraphs as you type. It also shows how many text messages your words take and how close you are to the limits on X, Instagram and in search results.
As a text message (SMS)
Type something to see how many SMS segments it takes.
Platform limits
Most used words
Words used more than once appear here, leaving out common words like "the".
0 bytes as UTF-8.
Runs entirely in your browser. Nothing you type or upload is sent to a server.
What counts as a character?
It depends who is counting, which is why two tools can disagree about the same text. A person sees a character as one symbol on the screen. Software can count Unicode code points, UTF-16 code units (what JavaScript's .length returns), or bytes of storage. For plain English these are all the same number. For accented letters and emoji they are not:
| Text | What you see | Code points | UTF-16 units | UTF-8 bytes |
|---|---|---|---|---|
| A | 1 | 1 | 1 | 1 |
| é | 1 | 1 or 2 | 1 or 2 | 2 or 3 |
| 👍🏽 | 1 | 2 | 4 | 8 |
| 🇺🇸 | 1 | 2 | 4 | 8 |
| 👨👩👧 | 1 | 5 | 8 | 18 |
The "é" row has two answers because it can be stored either as one precomposed letter or as a plain e followed by a combining accent. Both look identical. The family emoji is three people joined by invisible zero-width joiners. This counter reports what you see, using the same grouping rules your browser uses to move the cursor, and reports bytes separately for when you need to fit a database field or an API limit.
How text messages are counted
An SMS carries 140 bytes. In the standard GSM-7 alphabet each character takes 7 bits, so 160 characters fit. A few characters from GSM's extension table, such as { } [ ] ~ \ | and €, take two slots each. If even one character falls outside the alphabet, the phone switches the entire message to UCS-2, where each character takes 16 bits and only 70 fit.
Longer messages are split into segments and reassembled on the receiving phone. Each segment gives up 6 bytes to a header that says how to put them back together, which is why the per-segment limits drop to 153 and 67. Carriers and messaging services bill per segment, so one stray character can double or triple the cost of a bulk send.
The usual culprit is punctuation you never typed. Phones and word processors replace straight quotes and apostrophes with curly ones automatically, and the curly ones are not in GSM-7. The SMS panel names every character that forced Unicode so you can swap it out. There is a fuller explanation in Why one emoji cuts your text message limit in half.
Limits on social platforms and in search
Each platform counts differently. X counts any link as 23 characters, and counts emoji and Chinese, Japanese and Korean characters as 2, so a post can hit 280 well before the plain count does. Bluesky counts what you see, like this tool. Google does not cut search titles at a fixed number of characters at all; it truncates by width, at roughly 600 pixels, so a title of wide capital letters gets cut sooner than one of narrow lowercase ones. The 60 and 160 guides in the panel are safe targets rather than hard rules.
Reading and speaking time
Reading time assumes 238 words per minute, the average silent reading speed for non-fiction in Marc Brysbaert's 2019 review of reading-rate research. Technical material is read more slowly and light fiction faster. Speaking time assumes 140 words per minute, a comfortable pace for a talk, a podcast or a video script. For a timed speech, read it aloud once and adjust, because pauses and slides add time no word count can see.
Using the word list
The most used words panel lists words that appear more than once, leaving out short and common ones like "the" and "with". It is a quick way to spot a word you have leaned on too often in an essay, or to check that a page is actually about the topic you meant it to be about.
