What it does
Glitch text adds Unicode combining marks (accents, dots, tildes, hooks) above and below each letter. Combining marks are meant to attach to the character before them, so stacking several produces a distorted, glitchy look. This tool limits the stack to keep it legible. Each mark counts as one character on X, so a glitched word costs more than the plain one.
How do combining marks work?
A combining mark is a character with no width of its own that draws on top of, or under, the character before it.
Unicode uses them so that any letter can carry any accent: e followed by the combining acute accent (U+0301) displays as é. There's no rule against adding a second or third mark, and fonts will try to stack them. Glitch text exploits that. The tool picks marks from the Combining Diacritical Marks block, split into ones that sit above the letter and ones that hang below.
| Strength | Marks above | Marks below |
|---|---|---|
| Light | 1 | 0 |
| Medium | 1 to 2 | 0 to 1 |
| Strong | 2 to 3 | 1 to 2 |
How much of my post does glitch text use?
Every mark is a separate character, and X counts each one as one.
The marks come from a range X weighs as a single character, so a five-letter word at medium strength usually costs between ten and fifteen characters instead of five. At strong, it can triple. The counter on each result shows the exact number, and the X character counter will highlight what's eating your space.
Is strikethrough text the same thing?
Yes, technically. Strikethrough adds the combining long stroke (U+0336) after every character, and underline adds the combining low line (U+0332).
That's why strikethrough text works in places that don't support formatting, like X posts, and also why it doubles the character count. It's handy for jokes ("I ~~love~~ tolerate Mondays") but shouldn't carry information, since screen readers usually ignore the stroke entirely.
Questions people ask
Why not allow more marks?
Heavy zalgo text overlaps neighbouring lines, makes posts hard to tap and select, and is miserable for screen reader users, who may hear every single mark read out. A light effect gets the look with fewer side effects.
What does "Shuffle the marks" do?
It picks a different random set of marks. The randomness is seeded, so the same text and setting give the same result until you shuffle again.
How do I remove glitch marks from text?
Paste it into the decoder, which strips loose combining marks. Precomposed accented letters such as é are kept.