Zero Width Space Detector — Find Hidden Zero-Width Characters Online

A free, browser-based zero-width space detector for writers, editors, and developers who need to find and remove invisible U+200B characters — no sign-up and nothing uploaded.

Detect Hidden Zero-Width Spaces

Drop a plain-text file here to check and clean, or click to browse

.txt, .md, .csv, .json, .html, .css, .js, .ts, .py, .java, .go, .xml, .log

▶ Scan rules (4 groups enabled)

What Is a Zero-Width Space?

A zero-width space (ZWSP, code point U+200B) is a Unicode character that renders as nothing at all. It takes up no visible width, and unlike a normal space it leaves no gap between the characters around it. To a reader, the two halves of a word split by a zero-width space look completely joined. To a computer, they are still two separate tokens.

That double nature is what makes the character notable. Browsers and word processors can use a zero-width space to hint where a long word is allowed to wrap, so it has legitimate typographic uses. At the same time, because it is invisible, it can sit inside a word, a URL, an email address, or a line of code without anyone noticing — which is how it ends up breaking search, hitching a ride through copy-paste, or carrying a hidden mark.

This page is a dedicated zero-width space detector. Its job is not to rewrite your text but to show you exactly where these invisible characters are hiding, in whatever you paste or upload.

Which Zero-Width Characters This Detector Finds

The scanner flags the whole zero-width family, not just U+200B, because the same "invisible but meaningful" problem comes from several code points:

  • U+200B Zero Width Space — the classic invisible break; the one most often found in copied text.
  • U+200C Zero Width Non-Joiner — stops two characters from forming a joined, ligature-style shape.
  • U+200D Zero Width Joiner — asks for a joined form; widely used inside emoji sequences.
  • U+2060 Word Joiner — tells the renderer not to allow a line break at its position.
  • U+FEFF Zero Width No-Break Space / BOM — a byte-order mark that sometimes appears at the very start of a file or string.
  • U+180E Mongolian Vowel Separator — deprecated in modern Unicode but still present in older data.

Each match is highlighted in place and listed with its code point, so you can tell a harmless design choice from something that was inserted on purpose.

Why a Hidden Zero-Width Space Causes Problems

A character you cannot see is a character you cannot check by eye, and that is where most of the trouble starts.

  • Search and filtering break. Text copied from a web page, a PDF, or a chat window often carries zero-width spaces. A search for invoice will not match a string that looks identical but actually holds a zero-width space in the middle of the word.
  • Comparisons and deduplication fail. Two strings that render the same can differ byte for byte, so sorting, hashing, and "find duplicates" routines treat them as different values.
  • Form validation misbehaves. A pasted email, coupon code, or account number with an invisible character is rejected as invalid, and CSV or JSON imports can fail for the same reason.
  • Code will not compile. A zero-width space dropped into a variable name or a file path is not a legal character in most languages, yet it is invisible in the editor.
  • Text can carry a hidden mark. A short run of zero-width characters can encode a message or a fingerprint that survives copy-paste while staying invisible to the reader.

How to Detect Zero-Width Spaces

The scanner at the top of this page is built for exactly this job. To find hidden zero-width characters:

  1. Paste your text into the box, or drop a plain-text file onto the upload area. Everything stays in your browser.
  2. Run the scan. Every zero-width character that matches a rule is flagged with its name and code point.
  3. Read the highlighted view to see each hidden character in the exact position where it sits, surrounded by its real context.
  4. Open the findings report for the full list of matches, their severity, and the surrounding text.
  5. Copy the cleaned output from the Cleaned Text tab if you want those characters removed after you have reviewed them.

Everything happens locally. No sign-up, no upload, no data retention.

Detect First, Then Decide

Detection and cleaning are two different decisions, and it helps to keep them apart. Seeing what is there — and where — tells you whether a zero-width space was an accident, a formatting trick, or something inserted deliberately. Only once you know that can you sensibly choose to delete it.

If your goal is simply to strip every invisible character from a batch of text, the Unicode Text Cleaner is shaped around that workflow. This page is shaped around looking first.

Rule of thumb: treat any zero-width character inside a name, code, address, or identifier as suspicious until you have confirmed why it is there. In ordinary prose, its most likely origin is copy-paste from another app.

FAQ

How is a zero-width space different from a normal space?

A normal space has a visible width and opens a gap between characters. A zero-width space has no width at all: it inserts a character boundary with nothing visible on screen, so the text on either side stays pushed together.

Can I see a zero-width space in my text?

Not directly. It is invisible by definition, so no editor or preview shows it on its own. You need a scanner — like the one on this page — to reveal its position and code point.

Is my text uploaded anywhere when I scan it?

No. The detector runs entirely in your browser. Pasting or uploading text does not send it to a server, and nothing is stored or logged.

Can this tool also remove the zero-width spaces it finds?

Yes. After you review the findings, the Cleaned Text tab lets you copy the same text with those characters stripped out. For a broader cleanup that also handles unusual spaces and bidirectional controls, use the Unicode Text Cleaner.

Where does a zero-width space come from, and how does it get into text?

Most zero-width spaces arrive through copy-paste rather than being typed on purpose. They come from web pages that insert them as line-break hints, from PDF and office-document extraction, from chat clients and CMS templates, and once they reach the clipboard they travel with the rest of the text. Deliberately typing one is rare outside typography and obfuscation.

How can I check for zero-width spaces manually without a scanner?

There is no reliable way to spot one by eye, so you have to inspect the raw characters. Common methods are opening the file in an editor that shows code points, searching a hex view for the UTF-8 byte sequence E2 80 8B, or stepping through the text in a Unicode-aware debugger — all slow and easy to miss. This scanner does the same code-point comparison instantly in your browser, but it only reads character-level evidence and cannot detect statistical watermarks such as SynthID.

What is the difference between U+200B, U+200C, and U+200D?

U+200B is an invisible break that affects line wrapping only, while U+200C and U+200D control whether neighbouring characters form a joined shape. U+200C (zero-width non-joiner) blocks a ligature or cursive connection, and U+200D (zero-width joiner) requests one — it is also what holds multi-part emoji sequences together. The three are not interchangeable, and deleting a ZWJ by accident can split an emoji into its separate parts.

Why do search engines, forms, and source code break over an invisible character?

They break because a string that looks identical is not the same string once a zero-width character sits inside it. Search engines index the extra code point as part of the term, form validators reject a pasted email or code that contains it, and compilers refuse an identifier with characters outside the allowed set — and in every case the cause stays invisible on screen. This detector covers character-level traces such as U+200B, U+200C, U+200D, U+2060, U+FEFF, and bidirectional controls, but a hit is evidence of manipulation, not proof that the text was AI-generated.

Can zero-width spaces be removed in bulk, and will that change my layout?

Yes, every flagged zero-width character can be stripped from the whole text in a single pass. The visible words do not change, but the invisible break points a designer may have placed for line wrapping disappear, so long URLs or compound words can wrap in a slightly different place. If the text is typeset for print, review the positions before cleaning.

Can zero-width spaces be used to prevent copy-paste or track who reads a document?

Yes, that is exactly how they are used: a short run of zero-width characters can encode a fingerprint or a hidden message that survives copy-paste. The pattern is invisible to readers but can be read back programmatically, which is why the same trick appears in leak tracing, anti-copy schemes, and social-media tracking. This page only reveals those characters and their code points; it does not decode or attribute them.

Related Guides

← Back to AI Watermark Detector

Related Tools