👻 Invisible Character Detector & Remover
Reveal and strip the invisible characters hiding in pasted text — zero-width spaces, non-breaking spaces, and the tag characters increasingly used to secretly watermark AI-generated text.
⚠ 2 hidden characters found
- U+200B Zero Width Space ×1 zero-width
- U+00A0 No-Break Space ×1 space
Detects zero-width spaces, non-breaking spaces, bidirectional controls, the BOM, and the tag/variation-selector characters increasingly used to invisibly watermark AI-generated text. 🔒 Everything runs in your browser — your text is never uploaded.
How the invisible character detector & remover works
The tool walks your text code point by code point and flags every character that renders as nothing or as a deceptive space: zero-width spaces and joiners, the byte-order mark, bidirectional controls, non-breaking and exotic spaces, and the Unicode tag / variation-selector characters that some systems use to embed an invisible watermark. It lists what it found and gives you a cleaned copy with the hidden characters removed and odd spaces normalised.
This is exactly the kind of thing a chatbot cannot help with — an LLM literally cannot see or reliably emit invisible characters. It matters because pasted text from PDFs, websites and AI tools often carries hidden characters that break code, search, and form validation, or that fingerprint the text.
Frequently asked questions
What are invisible / hidden characters?
Unicode code points that take up no visible space or masquerade as a normal space — zero-width spaces (U+200B), the word joiner, non-breaking spaces, the byte-order mark (U+FEFF), bidirectional controls, and tag characters. They are invisible on screen but very real in the bytes.
Does AI-generated text contain hidden watermarks?
It can. Some systems embed invisible tag or variation-selector characters to fingerprint AI output. This tool flags those ranges specifically, so you can see and remove them.
Why does hidden text break things?
A zero-width space inside a keyword stops a search or an if-statement from matching; a non-breaking space can break CSV parsing or number formatting; bidi overrides can make a filename display deceptively. Removing them makes text behave predictably.
What does the cleaner do to spaces?
Exotic spaces (non-breaking, narrow no-break, ideographic) are converted to a normal space, line/paragraph separators become newlines, and zero-width and control characters are dropped entirely.
Is my text uploaded?
No — the scan and clean run entirely in your browser, which matters because the text you are checking may be sensitive.
Related text tools
From the blog
-
Homoglyph Attacks: How Lookalike Characters Spoof Domains and Brands
“pаypal.com” with a Cyrillic а looks identical to the real thing but is a different address. How homoglyph (IDN homograph) attacks use confusable Unicode characters for phishing, why mixed scripts are the giveaway, and how to detect them.
-
Invisible Characters in Text: Zero-Width Spaces, Hidden Watermarks, and How to Remove Them
A zero-width space (U+200B) is invisible on screen but real in the bytes — it breaks search, code and CSV parsing, and tag characters (U+E0000–E007F) can invisibly watermark AI text. What the hidden characters are, why they matter, and how to detect and strip them.