Embed Watermark in Text
Hide a short ID, tag or message inside any text — invisibly. The watermarked text looks identical to the original, but anyone with the scanner can prove it is yours.
Why Watermark Your Own Text?
Invisible watermarks turn ordinary text into a carrier of hidden proof:
- Prove authorship — embed your name or ID in drafts before sharing them; if the text resurfaces uncredited, the hidden payload shows where it came from.
- Trace leaks — send each recipient a copy tagged with a unique invisible ID. When a document leaks, decode the watermark and you know whose copy it was.
- Tag versions — mark drafts with a hidden version string so forwarded copies stay identifiable.
- Seed datasets — publishers hide markers in their text to detect unauthorized scraping and reuse.
How the Embedding Works
Your payload is encoded as UTF-8, each byte becomes 8 bits, and every bit becomes an invisible character — zero-width space for 0, zero-width non-joiner for 1. The bit stream is wrapped between two invisible delimiter characters (U+2062 and U+2063) and interleaved between the words of your text, one hidden character per gap where possible, so the watermark spreads through the whole document instead of sitting at the end.
The result is visually identical to your original text. Verify it any time with the watermark scanner.
Good Practices
- Keep the payload short — an ID or tag, not a paragraph.
- Always keep an unwatermarked original; you cannot "un-embed" selectively.
- Test the watermarked copy with the scanner before distributing it.
- Copy the watermarked text as-is — retyping or editing by hand can destroy the hidden stream.
- Use it for provenance, not deception — watermarking text you did not write to impersonate someone else is a misuse of the technique.
Frequently Asked Questions
Why would I embed a watermark in my own text?
The classic uses: proving authorship of your writing, tracing leaks (give each recipient a copy with a unique hidden ID and the leaker reveals themselves), and tagging document versions so you can tell drafts apart after they have been forwarded around.
How much text can I hide?
The payload is limited to 64 characters — enough for an ID, email, version tag or short message. Each character becomes 8 invisible bit-characters, so short payloads keep the watermark compact and robust.
Can someone remove or forge my watermark?
Yes — invisible-character watermarks are fragile by nature: anyone who knows the scheme can strip or rewrite it (this very site offers a remover). Treat it as a lightweight provenance tag, not cryptography. For legal proof of authorship you need signatures and timestamps, not hidden characters.
Will the watermark survive copy-paste?
Yes, as long as the text is copied as text — the invisible characters travel with it. It is lost if someone retypes the text by hand, screenshots it, or runs it through a cleaner like our Remove tab.