Zero-width space remover

A zero-width space, U+200B, occupies a position in a string of text and produces no visible output whatsoever. It sits in the byte stream, invisible on screen, present in every comparison, search and parse performed against the text. This tool detects and removes it along with a small set of related zero-width characters, without altering any character that actually renders.

Detection and removal both run client-side, in JavaScript. There is no upload, no server round trip and therefore no run limit, since nothing here has a cost to meter.

There's nothing to preview beforehand, since the whole category of problem is invisible until the scan runs.

Step 1

Paste the text into the input field.

Step 2

Run the removal. It executes instantly, locally.

Step 3

Read the result. Every rendered character is unchanged.

U+200B is the character most commonly meant by 'zero-width space', but it rarely appears in isolation. A small set of related codepoints share the property of occupying a position in the text while rendering nothing.

U+200B zero-width space

The baseline case: a spacing character with no visual output. The most frequent invisible marker in copied text.

U+200C zero-width non-joiner

Prevents adjacent letters from visually joining. Functional typography in Persian, Arabic and several Indic scripts.

U+200D zero-width joiner

Fuses adjacent characters, or emoji, into a single displayed glyph.

U+2060 word joiner

Marks two characters as unbreakable across a line wrap, with no other visible effect.

Not every instance of these characters is an artifact to be removed. A zero-width joiner sitting between two emoji is precisely what composes them into a single combined glyph, a family emoji or a profession emoji built from two source characters. Removing that joiner would visually split the composed emoji apart, so a joiner detected inside a genuine emoji sequence is left in place by this tool.

The zero-width non-joiner carries equivalent weight in Persian and several Indic scripts, where it interrupts a script's default joining behavior to keep two letters visually distinct. Text in those languages should have its zero-width non-joiner instances reviewed individually rather than stripped as a batch, since what reads as noise in Latin-script text can be functional punctuation there.

Inside a heading or a URL slug, a zero-width space splits the string for any process comparing it byte by byte: full-text search, a regular expression, a string equality check. Two values that render identically on screen fail to match, because one carries a codepoint the other does not.

In source code, a zero-width character embedded in an identifier or a string literal produces an error with no visible cause in most editors. In tabular data, the same hidden character inside a field can silently fail a lookup or a join that should have succeeded.

Engineers cleaning text pulled from documentation, chat exports or PDFs before it enters a codebase, a config file or a commit run it here first, since the resulting failure mode is otherwise a syntax error with no visible trace of its cause. Publishing and SEO teams use it when a slug or on-page search stopped matching a heading for no apparent reason.

It is standard practice in any workflow performing programmatic comparison: deduplication, data validation, database imports, anywhere two values that should be logically identical need to actually match byte for byte.