/SHRUG EMOJI · EMOJI COMBOS · TEXT FACES

Shrug Emoji ¯\_(ツ)_/¯, Emoji Combos and 122 Text Faces to Copy

The shrug emoji is right here. Press Copy the shrug to take the one on show, or click any variant below to copy that one straight away.

Nine characters, no image. It pastes into a commit message, a DM or a terminal and still looks like a shrug. Clicking a variant below copies it and loads it up here.

Variants, click one to copy it

It is also called the shrug emoticon, and it is technically a kaomoji, because it is built from ordinary characters rather than a single emoji codepoint. Below it is a searchable library of 122 text faces and cute kaomoji sorted by mood: the lenny face, table flip ascii, ascii cat, happy, sad, angry, love, animals, dance and confused. Under that sit 118 emoji combos for bios and captions. Everything runs in your browser, nothing is uploaded.

/LIBRARY

Browse every text face and kaomoji

Type to filter by name, mood or the characters themselves, or pick a category. Clicking a face copies it straight away, and it also loads into the preview above the grid so you can see what you took. The shrug emoticon, the lenny face, table flip ascii and the ascii cat all live here.

Every face here is plain text. If you want the multi-row kind of picture instead, the text to ASCII generator turns a word into a large banner, and the glitch tool turns a photo into a broken-signal render.

/EMOJI COMBOS

Emoji combos to copy and paste

An emoji combo is a short run of two to five emoji that reads as one idea: a mood, an aesthetic, a caption. These are the ones people actually paste into Instagram and TikTok bios, display names, story text and Discord statuses. Search by name or by the emoji themselves, or pick a category. Clicking a combo copies it straight away, and it also loads into the preview above the grid so you can see what you took.

Combos are real emoji, so each platform draws its own artwork and the same combo looks a little different on an iPhone, on Android and on the web. For something that renders identically everywhere, use a text face instead.

/WHAT THEY ARE

What text faces and kaomoji actually are

A text face is a face drawn with characters your fonts already have. Nothing is uploaded, nothing is rendered, nothing is an image. The shrug is nine characters sitting in a row, and every one of them is an ordinary Unicode codepoint that any modern text field will accept.

Four things get confused with each other, and the difference matters the moment you paste.

Emoticons are the Western sideways kind, :-) and ;) and :'(. You read them by tilting your head. They came out of 1980s bulletin boards, where a plain colon and a parenthesis were the whole toolkit.

Kaomoji are the Japanese kind, and kao-moji literally means face-characters. They read upright, like (^▽^), with the eyes doing most of the work and the mouth doing almost none. They grew on Japanese message boards through the 1980s and 1990s, where katakana, full-width punctuation and box-drawing characters gave people a far bigger drawing kit than a US keyboard offered. That is why sits in the middle of the shrug: it is the katakana letter tsu, borrowed purely because it looks like a smirk.

Emoji are single codepoints that a font renders as a small picture. A smiley emoji is not built from parts, it is one character your device draws for you. That makes it compact and colourful, and it also means you get whichever picture the platform decided on, which is why the same emoji can read as friendly on one phone and unhinged on another.

Emoji combos are the newest of the group and the only one built out of emoji rather than characters. A combo is two to five emoji placed in a deliberate order so the run reads as one idea: 🌟🌙💫 is a mood, not three separate pictures. Combos took over bios and captions because a bio field gives you very little room, and three emoji set a tone faster than a sentence does. The trade is the same one emoji always makes: the artwork belongs to the platform, so your combo is redrawn on every device it lands on.

ASCII art is the fourth cousin: a whole picture assembled from a grid of characters, dozens or hundreds of rows deep, which needs a monospaced font to stay aligned. Text faces are one row tall, so they survive in places ASCII art cannot. If you want the multi-row kind, that is what an ASCII art generator is for.

/WHY THEY LAST

Why text faces outlive the images around them

A text face is text. That single fact is the whole reason people still paste the shrug forty years after the first emoticon appeared.

Text goes everywhere. It survives a copy into a commit message, a ticket title, a git branch description, a terminal, a plain-text email, a CSV export, a printed page. There is no upload step, no CDN, no hotlink to rot, no image that fails to load on a slow connection. The character sequence is the artwork, so wherever the characters go, the art goes with them.

Text also carries tone in a way a standardised picture cannot. Emoji are designed for legibility, which makes them generic. A kaomoji is assembled, so the exact combination you pick reads as a specific voice. The classic shrug and the casual ╮( ̄▽ ̄)╭ variant in the library above both mean "I do not know", and they do not mean it the same way. One is resigned, one is breezy, and the difference is entirely in which characters you chose.

And text faces are stable in a way images are not. Emoji artwork gets redrawn every few years, so a message you sent in 2016 can read differently today without you touching it. The shrug's macrons and katakana render the same now as they did the day they were first posted, because nobody owns them and no design team is redrawing them next season.

/ANATOMY

What the shrug emoticon is made of

The shrug is nine characters, and every one of them is doing a job. If you ever need to rebuild it by hand, or work out why a paste came out wrong, this is the map.

#CharacterCodepointWhat it isWhat it draws
1¯U+00AFMacronLeft hand, palm up
2\U+005CReverse solidus, the backslashLeft arm
3_U+005FLow lineLeft shoulder
4(U+0028Left parenthesisSide of the head
5U+30C4Katakana letter tsuThe face and the smirk
6)U+0029Right parenthesisOther side of the head
7_U+005FLow lineRight shoulder
8/U+002FSolidus, the forward slashRight arm
9¯U+00AFMacronRight hand, palm up

Two things are worth remembering from that table. The backslash is the only character in the set that any software treats as special, which is exactly why it is the one that goes missing. And the macron is neither a hyphen nor an overline: swapping it for either drops the hands to the wrong height and the shrug stops reading as a shrug.

/WHERE THEY BREAK

Where text faces break, and how to fix it

Four failures cover almost every report of a face arriving wrong. Three of them have a fix, and the fourth has a workaround.

  1. The Markdown parser ate your backslash

    Reddit, Discord, GitHub and several Slack contexts parse Markdown, where a backslash escapes the character after it. Paste the shrug into one of those and it can arrive one arm short.

    FixDouble it, so the parser prints a single backslash. Or wrap the whole face in backticks, which stops anything inside from being parsed at all.

  2. The font has no glyph for it

    Stripped-down Linux images and older Windows installs ship without Japanese coverage, so katakana and the rarer symbol blocks fall back to an empty box. There is nothing the reader can do about it.

    FixFor audiences you cannot control, prefer faces built from Latin, Greek and box-drawing characters. (¬_¬) travels further than anything with a bear snout in it.

  3. Combining marks land in the wrong place

    Faces like the lenny stack an accent onto the character before it. Some renderers, and almost every text field with aggressive line-height clipping, put the mark somewhere unhelpful and the eyebrows drift off the face.

    FixPreview it once in the destination before you send. If it looks wrong there, take a variant with no combining characters, which is most of the library above.

  4. A validated field strips it

    Username fields, SMS gateways and older databases often accept a limited character range, and they either reject the face outright or store it as a row of question marks.

    FixAssume anything with a length limit measured in bytes will mangle it. Keep the face in the message body, out of fields that get validated on the way in.

/WHERE THEY LAND

Where text faces work best

Chat and DMs

Slack, Discord, iMessage, WhatsApp. If a backslash keeps vanishing in Slack or Discord, wrap the face in backticks and it arrives intact.

Commits and tickets

Git commit messages, pull request descriptions, issue trackers. Plain text end to end, so nothing is stripped and nothing needs an attachment.

Social posts

X, Reddit, Mastodon, forums. Reddit renders Markdown, so escape the backslash in the shrug or the post will show it with one arm.

Terminals and code

Log lines, CLI splash text, code comments and docstrings. Any font with the right coverage shows them, and they cost nine bytes rather than an image request.

/FAQS

Common questions about text faces

How do I copy the shrug emoji without it losing an arm?
Copy it from the top of this page, then double the backslash or wrap the whole face in backticks wherever you paste it. The shrug is nine characters and one of them, the backslash, is an escape character in Markdown, so Reddit, Discord, GitHub and some Slack contexts swallow it and the left arm disappears. Backticks stop anything inside being parsed at all.
Why does my shrug emoji show up as a box or a question mark?
Because the font on the reader's device has no glyph for one of the characters, almost always the katakana tsu in the middle. Stripped-down Linux images and older Windows installs ship without Japanese coverage, and there is nothing the reader can do about it. For an audience you cannot control, pick a face built only from Latin, Greek and box-drawing characters.
How do I set up a shortcut so I can type the shrug instead of pasting it?
Save it as a text replacement on your device: copy the shrug from the top of this page, then add it under Settings, Keyboard, Text Replacement on iPhone and Mac, or as a personal dictionary entry on Android. Almost nobody types it by hand, because the macron and the katakana tsu are not on a standard keyboard.
Is the shrug emoji actually an emoji?
No, strictly it is an emoticon, and more precisely a kaomoji, because it is assembled from nine ordinary characters rather than being one emoji codepoint. Most people search for it as the shrug emoji anyway, since emoji has become the everyday word for any small expressive symbol. It is the same nine characters under either name.
Which emoji combo should I put in my Instagram or TikTok bio?
Pick one that sets the tone in two or three emoji, because bio fields are short and many apps count a single emoji as two characters. The combos on this page are sorted into 12 moods, aesthetic, love, cute, party and so on, so you can choose by feeling rather than scrolling. Put it on its own line or at the end of a line, since some apps wrap mid-run and split a combo in two.
Why do emoji combos look different on my friend's phone?
Because every platform draws its own emoji artwork, so the same codepoints become different pictures on an iPhone, on Android and on the web. The characters and their order never change, only the drawings do. If you need something that renders identically everywhere, use a text face instead, because a text face is drawn by the font the reader already has.
How do I make the Lenny face on my keyboard?
You do not type it, you copy it, because it needs an inverted glottal stop for the mouth and two combining breve marks that no keyboard offers. Copy one of the eleven Lenny variants from the library above and save it as a text replacement if you use it often. The two degree signs are the eyes, and the breves above and below are the eyebrows and the smirk.
How do I do the table flip in a chat message?
Copy it from the Table flip category in the library above, which holds ten variants including the calm one that puts the table back down. It is a figure with raised arms, a wave mark, and a table drawn from box-drawing characters, so it needs nothing beyond a normal Unicode font. Wrap it in backticks on Discord or Reddit so the parser leaves it alone.
Can I put a text face in a GitHub README without breaking the Markdown?
Yes, wrap it in backticks and it comes through untouched. Without them GitHub's Markdown parser reads a backslash as an escape, so the shrug arrives one arm short, and combining marks can drift off the face. Commit messages need no escaping at all, because git stores plain text and never parses it.
Do text faces work on iPhone and Android?
Yes. Both ship fonts with wide Unicode coverage, so katakana, box drawing, Greek letters and combining marks all render properly. The faces most likely to fall back to an empty box are the ones using rarer symbol blocks, and that happens on stripped-down desktop font installs rather than on modern phones.
What is the difference between a kaomoji and a Western emoticon?
A kaomoji is read upright and a Western emoticon is read sideways. A colon and a bracket needs you to tilt your head, a kaomoji does not. Kaomoji grew on Japanese message boards where katakana, full-width punctuation and box-drawing characters offered a far bigger drawing kit than a US keyboard, which is why they carry so much more expression per character.
How do I make an ASCII cat or a cat emoticon?
Copy one from the Animals category in the library above, which holds five cats: the classic cat emoticon with whiskers, a grumpy one, a waving one, a small sideways face and a cat peeking over an edge. An ASCII cat is a text face like any other, so it pastes into a bio, a commit message or a Discord line with no image and no upload.
What do I use when I need a multi-row ASCII picture rather than a one-line face?
Use an ASCII art generator, because a text face is one row tall by design. The text to ASCII generator turns a word into a large banner assembled from characters, and the ASCII Magic studio turns a photo or a video into a full character picture. Both need a monospaced font to stay aligned, which is exactly the constraint a one-row text face escapes.
Where do text faces still work when emoji and images do not?
Anywhere that stores plain text and refuses pictures: git commit messages, branch descriptions, issue titles, terminal output, log lines, code comments, plain-text email and CSV exports. The character sequence is the artwork, so there is no upload, no CDN and no image that fails to load on a slow connection.

/DO MORE WITH ASCII MAGIC

One row of characters is the small version.

The studio behind this page turns photos and video into full ASCII art, pixel art, dither, halftone, voxel and more. Export PNG, GIF or MP4, all in the browser.

Open the editor →

/RELATED

More ways to make things out of characters

This page is for one-line faces. To turn a word into a large banner, use the text to ASCII generator. To turn a photo into a broken-signal render, use the glitch tool. For everything else, open the full studio.