One Emoji, Four Character Counts: What Unicode 18 Changes and What It Doesn't

On September 16, 2026, Unicode 18.0 was released. According to Unicode's version page it adds 13,007 characters, for a total of 172,808, including three new scripts: Proto-Cuneiform numerals, Jurchen and Seal. Emojipedia counts nine new emoji: Cracking Face, Leftward Thumb Sign, Rightward Thumb Sign, Monarch Butterfly, Pickle, Lighthouse, Meteor, Eraser and Net With Handle. With skin tone variants for the two thumb gestures, that makes 19 additions. This post is not about a limit. It is about a quieter question: when a tool says your text has 200 characters, what exactly did it count?

Want to see how a piece of text counts? Clear Count counts words and characters in your browser, and nothing is uploaded.

Released Is Not the Same as Usable

Emojipedia is explicit that the release "does not mean users can immediately access or use any new emoji." Its estimates are late 2026 for some early adopters, Q1 2027 for Android, Samsung and WhatsApp on Android, March or April 2027 for Apple, and late summer or fall 2027 for Microsoft. These are estimates, not announced dates. Until then, the same message can display differently on a sender's phone and a recipient's, which matters for counting.

One Emoji, Four Counts

I ran five emoji through Node 22 and counted them four ways. UTF-16 code units are what JavaScript's .length reports. Code points are what Python's len() returns. Graphemes are what a person would call "one character." UTF-8 bytes are what many databases and APIs store.

MDN documents the first column: length "contains the length of the string in UTF-16 code units," so characters outside the Basic Multilingual Plane, including most emoji, count as two. The family emoji shows how it compounds: four people joined by three invisible zero width joiners, which is also why it appears in our post on invisible characters in copied text.

Unicode's emoji specification says all emoji sequences are single grapheme clusters, with no boundary inside them, so the cursor and the delete key treat the family as one thing. That rule governs editing. It does not tell any platform to count an emoji as one against a limit, and each platform picks its own unit.

What the New Emoji Add

The thumb signs with skin tones follow the same two-part pattern as the 👍🏽 row: a base emoji plus a modifier, which is two code points and four UTF-16 units for one visible character. There is also a display issue. The specification says that when a system has no combined glyph for a joined sequence, "the ZWJ characters are ignored and a fallback sequence of separate emoji is displayed." During the 2027 rollout, a sender on a new phone may see one symbol while a recipient on an older one sees two.

What to Do About It

• Find out which unit the platform uses. If a limit matters, test with an actual emoji on the real platform. The SMS case, where one emoji changes the encoding and cuts the limit, is covered in SMS Character Limit.
• In code, pick the right method. Use [...str].length for code points and Intl.Segmenter with grapheme granularity for what people see. MDN marks Intl.Segmenter as Baseline since April 2024.
• Leave a margin. Where a limit is tight, assume an emoji may cost two to eleven units.

---

The honest summary: Unicode 18.0 added nine new emoji, and they will reach most phones during 2027. The more lasting point is that there's no single "character count" for an emoji. The same family emoji is 1, 7, 11 or 25 depending on whether a tool counts what you see, code points, UTF-16 units or bytes. Know which one your platform uses before an emoji decides whether a message fits.