What is a zero-width character and where does it come from?

A zero-width character is a real Unicode code point that takes up no visual space, so you cannot see it but it is in the bytes. The common one is U+200B ZERO WIDTH SPACE. They sneak in when you copy from chat apps, PDFs, Word, a CMS rich-text editor, or paste output from an AI model. Example: copy a function name from a styled doc and you may get "getUser" with a hidden U+200B between get and User — it looks fine but the identifier no longer matches.

Why does text I copied break my code or a JSON parse?

Because an invisible character came along for the ride. A leading U+FEFF (BOM) at the start of a JSON file makes JSON.parse throw "Unexpected token" even though the file looks perfect. A U+00A0 NO-BREAK SPACE where you expected a normal space makes "a b" !== "a b" so string comparisons and split(' ') quietly fail. Paste the text here, see the exact code points, strip them, and the parse works again.

What is the difference between U+200B and U+FEFF?

Both are zero-width and both render as nothing, but they have different jobs. U+200B ZERO WIDTH SPACE is a soft line-break opportunity used inside long words. U+FEFF is the byte-order mark: when it sits at the very start of a file it signals encoding, but anywhere else it acts as a zero-width no-break space. A stray U+FEFF mid-string is almost always junk from a bad export; a leading one breaks JSON and CSV parsers. This tool flags both by name so you know which you are looking at.

Will it delete my normal spaces by mistake?

No. A plain ASCII space (U+0020) and newlines are never flagged or touched. Only the invisible lookalikes are: NO-BREAK SPACE (U+00A0), THIN SPACE, IDEOGRAPHIC SPACE (U+3000) and the rest. In remove-all mode those get deleted; in normalize mode they become a normal space instead, so "price:\u00A09" turns into "price: 9" with word spacing intact rather than "price:9".

How do I detect zero-width characters used as an AI watermark?

Some systems hide a pattern of U+200B / U+200C / U+200D between visible letters to fingerprint or watermark generated text. Paste the suspect text and the detector lists every zero-width hit with its count and marks each position in the preview. If you see a run of zero-width joiners scattered through otherwise normal prose, that is the watermark — hit remove-all and copy the cleaned text to neutralize it.

Fix a JSON or CSV that won't parse

A config file throws "Unexpected token in JSON at position 0" but looks identical to a working one. The culprit is almost always a U+FEFF BOM at the front from a Windows export. Paste the file, see the BOM flagged at position 0, remove all, copy the clean version, and the parser stops choking.

Clean a function name or variable copied from docs

You copy "getUserById" out of a styled API doc into your editor and it errors as undefined. Hidden U+200B characters split the identifier. Paste it here, the table shows two zero-width spaces, hit remove-all, and paste back a name that actually matches the symbol.

Strip a zero-width watermark from AI or pasted text

Text generated by a model or lifted from a tracked document can carry a pattern of zero-width joiners as a covert fingerprint. The detector lists every U+200D hit and marks its position; remove-all neutralizes the pattern so the text is plain content again.

Normalize no-break spaces before a string comparison

A scraped price field reads "$ 9.99" but your equality check fails. The space is a U+00A0 NO-BREAK SPACE, not a real space. Switch to normalize mode to turn it into a plain space, copy the result, and your comparison and split(' ') start working.

Zero Width Character Detector: Find and Remove Invisible Characters

Find and strip zero-width spaces, the BOM, no-break spaces and other invisible characters, then copy clean text — runs in your browser

Runs locally
Category Text
Best for Removing repetitive cleanup work from everyday writing and operations.

Paste text to scan

Cleaning

Delete every invisible character.

What this tool does

Paste text and this tool finds every zero-width and invisible character hiding in it, then cleans it out. It catches the zero-width space (U+200B), zero-width non-joiner (U+200C), zero-width joiner (U+200D), the byte-order mark (U+FEFF), no-break space (U+00A0), left-to-right and right-to-left marks (U+200E / U+200F), word joiner (U+2060), Mongolian vowel separator (U+180E) and a long tail of format controls and exotic spaces. For each kind it shows the code point (U+XXXX), the Unicode name and how many times it appears, and a preview marks every hit in red so you can see exactly where it was. Clean two ways: remove all invisibles, or normalize — turn invisible spaces into a normal space while deleting the zero-width family. Then copy the cleaned text in one click. Everything runs locally in your browser; nothing you paste is uploaded.

Tool details

Input: Files + Text; The page exposes text boxes, numeric controls, file pickers, or structured inputs depending on the tool.
Output: Live result + Copy + Preview; The result area focuses on usable output, with copy, download, or preview actions when supported.
Privacy: Browser-side processing; The main tool logic does not call an external API, so inputs normally stay in the current tab.
Save / share: Shareable URL state; Key settings are encoded in the URL so another person can reopen the same setup.
Performance budget: Initial JS <= 9 KB; No WASM budget is declared, keeping the tool quick to open on mobile.
Best fit: Text · Developer; Category and role tags drive related tools, internal links, and quick fit checks.

How to use

1. Input

Paste or drop your content into the tool panel.
2. Process

Click the button. All processing is local in your browser.
3. Copy / Download

Copy the result or download to disk in one click.

How Zero Width Character Detector fits into your work

Use it to clean, compare, reshape, or extract plain text before it goes into a document, CMS, spreadsheet, or prompt.

Text jobs

Removing repetitive cleanup work from everyday writing and operations.
Making text easier to compare, paste, publish, or feed into another tool.
Working with content locally when the text is private or unfinished.

Text checks

Scan for unintended whitespace, duplicate lines, and lost punctuation.
For long text, test the first few lines before applying the whole change.
Copy the final output only after checking the preview.

Good next steps

These links move the current task into a more complete workflow.

Real-world use cases

Fix a JSON or CSV that won't parse
A config file throws "Unexpected token in JSON at position 0" but looks identical to a working one. The culprit is almost always a U+FEFF BOM at the front from a Windows export. Paste the file, see the BOM flagged at position 0, remove all, copy the clean version, and the parser stops choking.
Clean a function name or variable copied from docs
You copy "getUserById" out of a styled API doc into your editor and it errors as undefined. Hidden U+200B characters split the identifier. Paste it here, the table shows two zero-width spaces, hit remove-all, and paste back a name that actually matches the symbol.
Strip a zero-width watermark from AI or pasted text
Text generated by a model or lifted from a tracked document can carry a pattern of zero-width joiners as a covert fingerprint. The detector lists every U+200D hit and marks its position; remove-all neutralizes the pattern so the text is plain content again.
Normalize no-break spaces before a string comparison
A scraped price field reads "$ 9.99" but your equality check fails. The space is a U+00A0 NO-BREAK SPACE, not a real space. Switch to normalize mode to turn it into a plain space, copy the result, and your comparison and split(' ') start working.

Common pitfalls

Assuming "looks identical" means "is identical". Two strings can render the same pixel-for-pixel yet differ by a hidden U+200B or U+00A0. Always check the bytes, not the glyphs, when an equality test fails on text that looks right.
Using remove-all on text that contains real emoji. The zero-width joiner U+200D glues multi-part emoji like 👨‍👩‍👧 together; deleting it splits the family into three separate emoji. The preview marks every U+200D so you can eyeball emoji before you clean.
Stripping the BOM from a file that another tool actually expects. A leading U+FEFF breaks JSON parsers but some legacy CSV importers rely on it to detect UTF-8. Know the consumer before you remove a leading BOM, or you trade one bug for another.

Privacy

Detection, cleaning and the preview are plain JavaScript that runs inside your browser tab. The text you paste never leaves the page, is never uploaded to a server, and is not logged anywhere. Only the cleaning mode (remove vs normalize) is kept in the URL so a shared link reopens with your chosen option; the text itself is deliberately never put in the URL, because pasted text here is often sensitive — leaked secrets, private documents, or AI output under review. Close the tab and nothing remains.

FAQ

Tool combos

Folks in your role tend to reach for these alongside this tool.

Developer

Browse all tools for this role

Zero Width Character Detector: Find and Remove Invisible Characters

What this tool does

Tool details

How to use

1. Input

2. Process

3. Copy / Download

How Zero Width Character Detector fits into your work

Text jobs

Text checks

Good next steps

Real-world use cases

Fix a JSON or CSV that won't parse

Clean a function name or variable copied from docs

Strip a zero-width watermark from AI or pasted text

Normalize no-break spaces before a string comparison

Common pitfalls

Privacy

FAQ

Unicode Character Inspector

Text File Cleaner

Find and Replace Text

String Escape / Unescape

Chinese Acupoint Locator

Add Line Numbers

Aesthetic Text Generator

AI Eval Planner

AI Model Comparison

AI Token Counter

Arabic 500 Essential Words

ASCII Art Generator