How long should an SEO title be, why does one accented letter split an SMS, and why can a single emoji count as 2 or even 11? Character counting for search results, text messages, social posts and technical fields, with examples.
A character limit is the maximum number of characters a field accepts or displays, and not every system counts those characters the same way. The short version: Google truncates titles by pixel width rather than a fixed count, so desktop results typically show around 50–60 characters, and meta descriptions are usually cut somewhere near 150–160. A single SMS carries 160 characters in the standard GSM alphabet but only 70 once it switches to Unicode. An emoji looks like one character to you, yet depending on the system it may count as 2, 4 or even 11 units.
To measure your text, the character counter reports characters, characters without spaces, words and lines; it counts emoji and combined characters as one, and your text never leaves your device. The slug generator turns a headline into a clean URL, the URL encoder shows how accented letters expand inside an address, the text diff checker compares two headline drafts line by line, the QR code generator shows how much text fits in a code, and the password generator produces passwords at the length a site allows.
“Character” can mean at least four different things: the letter or symbol the eye sees (a grapheme), a Unicode code point, a 16-bit unit as used by JavaScript and SMS, and a byte written to disk. For plain English letters all four agree. For accented letters, symbols and emoji they drift apart, and the same text can fit on one platform and overflow on another depending on which of them is being counted.
| Example | Visible | Code points | 16-bit units | UTF-8 bytes |
|---|---|---|---|---|
| a | 1 | 1 | 1 | 1 |
| é | 1 | 1 | 1 | 2 |
| € | 1 | 1 | 1 | 3 |
| 😀 | 1 | 1 | 2 | 4 |
| 👍🏽 (with skin tone) | 1 | 2 | 4 | 8 |
| 🇬🇧 (flag) | 1 | 2 | 4 | 8 |
| 👨👩👧 (family) | 1 | 5 | 8 | 18 |
The family emoji is really three person emoji glued together by invisible zero-width joiners, which is why a four-person family reaches 11 sixteen-bit units. Flags are pairs of special regional-indicator symbols spelling a two-letter country code. The é above is the precomposed form; the same letter can also be stored as a plain e followed by a combining accent, which turns one visible character into two code points.
The counter follows what the eye sees: an emoji, a skin-toned emoji or a flag each count as one. That is the measure closest to a reader's experience and the right starting point for titles, descriptions and bios. SMS and some databases, however, count 16-bit units or bytes, so for text containing emoji the counter's figure can be lower than what those systems will charge you. The sections below explain what each field actually counts.
Titles and descriptions are the first two things people read in search results. Google publishes no official length for either; it trims what it shows to fit the device width and may rewrite the text altogether. The figures below are therefore common truncation points, not rules.
| Field | Common target | Truncated by | Note |
|---|---|---|---|
| Title tag | 50–60 characters | Roughly 600 pixels on desktop | Put key words first |
| Meta description | 140–160 characters | Screen width; earlier on mobile | Google may pick its own snippet |
| URL slug | 3–5 meaningful words | Shortened in results | Use hyphens, not underscores |
| Image alt text | Short, one sentence | No official limit | Screen readers read all of it |
Because truncation is pixel-based, wide letters (M, W) and narrow ones (i, l) take up different space, and a 50-character title in capitals can be cut earlier than a 60-character one in lowercase. A practical habit: place the main phrase within the first 40 characters and the brand at the end, so the meaning survives even if the tail is clipped. When you are torn between two drafts, putting them one under the other and comparing the differences shows which words are really pulling their weight.
The description does not directly affect ranking, but it does affect whether people click. Write one or two sentences of around 150 characters that say how the page answers the query. If Google finds a passage that matches the search better, it may show that instead, so treat the description as a strong suggestion rather than a guaranteed shop window.
Short, readable, hyphen-separated addresses tell both people and search engines what a page is about. Accented letters add a twist: an “é” in an address becomes %C3%A9 when copied, six characters instead of one. Run “Café Menu Ideas” through the slug generator and you get cafe-menu-ideas, which sidesteps the issue. As a technical ceiling, the sitemap protocol requires URLs shorter than 2,048 characters, though any address approaching that is unreadable anyway.
Text messages use a 7-bit alphabet known as GSM 03.38. A single message has a 140-byte payload, which works out to 160 characters in that alphabet. If the text contains even one character outside it, the whole message switches to UCS-2 (16-bit) encoding and the limit drops to 70. Longer messages are split into segments, and because each segment carries reassembly information, capacity per segment shrinks.
| Encoding | Single message | Per segment when split | Example: 200-character text |
|---|---|---|---|
| GSM 7-bit | 160 characters | 153 characters | 2 segments |
| UCS-2 (Unicode) | 70 characters | 67 characters | 3 segments |
Some characters belong to an “extension table” and take two slots even in 7-bit mode, including € [ ] { } ~ ^ | \. For anyone sending bulk SMS that detail shows up on the invoice: a 161-character message is billed as two.
The GSM alphabet includes some accented letters, such as é, è, ü and ñ, but not others, such as á, í, ó and ú. A single curly apostrophe pasted from a word processor also falls outside it. Any of these can silently switch a message to the 70-character mode. The 3GPP TS 23.038 standard defines national language tables for several languages that keep such letters in 7-bit mode, but you cannot assume every handset and sending gateway uses them, and some phones simply replace the letters with unaccented ones.
No 7-bit table contains emoji, so a single one forces Unicode mode. On top of that, most emoji take two 16-bit units, using two of the 70 slots. A skin-toned thumbs-up uses four and a family emoji eight. In a marketing message, one emoji can easily push the text into three segments.
Platforms change their limits from time to time, grant different allowances to paid tiers and weight some characters differently. Treat the values below as commonly cited figures and check the platform's current help pages before you publish.
| Platform and field | Commonly cited limit | Worth knowing |
|---|---|---|
| X (Twitter) post | 280 characters (standard account) | Emoji and some scripts are weighted; links count as a fixed length |
| Instagram caption | 2,200 characters | Only the first lines show in the feed |
| Instagram bio | 150 characters | Line breaks count too |
| YouTube video title | 100 characters | Shown much shorter in recommendations |
| YouTube description | 5,000 characters | The first lines appear before “more” |
| LinkedIn post | 3,000 characters | Long text is folded in the feed |
Whatever the platform, the limit that matters most is the visible part. Fitting the core message into the first 100–125 characters means it gets read even by people who never tap “see more”.
"😀".length returns 2 because it counts 16-bit units. When you need the visible count, segment graphemes with Intl.Segmenter.Almost everywhere, yes. SEO titles, SMS and social media limits all include spaces. The counter's “characters without spaces” figure is for cases such as some translation and writing assignments that exclude them.
One to the eye, and it varies by system. A simple emoji is 2 units in 16-bit systems, skin-toned and flag emoji are 4, and joined family emoji are 8 or more. When writing into a tight field, budget at least 2 characters per emoji.
There is no penalty; Google ends the visible title with an ellipsis or rewrites it. The risk is that an important word sits in the clipped part. If the key words come first, a few extra characters do no harm.
Roughly. In English, about 5 to 6 characters per word including the space is a reasonable rule of thumb, so 1,000 characters is around 170 words. For an exact figure, use the counter's word count.
Related reading: alt text and file name rules in image optimization for SEO, how much data sensibly fits in a code in the QR code usage guide, and why IPv6 addresses are so long in what is an IP address.
