Skip to content
Simple Unmark

SynthID Text · documented

Gemini Watermark Remover for Text

Google documents SynthID Text as its method for watermarking generated text, and it works by shaping token choices rather than inserting a symbol. Simple Unmark rewrites Gemini text to replace that token-choice pattern, and removes any hidden characters the copy picked up along the way.

Original

Local scan runs as you type.

Hidden-character cleanup + wording rewrite

Google Gemini: current status

Last verified · first-party documented

Google documents SynthID Text as its method for watermarking AI-generated text.

Status
In production
Mechanism
A logits processor applied during generation. Google describes it as augmenting the model's logits with a pseudorandom g-function, so the watermark lives in which tokens were chosen rather than in any added symbol.
Hidden characters involved?
No. SynthID Text does not insert hidden characters; nothing is appended to the visible text.
Public detector
Google publishes the watermarking and detection approach and has open-sourced a reference implementation, but a public detector for arbitrary Gemini text is not offered as a general consumer service.

What the provider says weakens the signal

  • Google states detector confidence "can be greatly reduced when an AI-generated text is thoroughly rewritten"
  • Translation into another language
  • Factual responses, where there is less opportunity to augment generation
  • Short excerpts, though the method is robust to cropping and mild paraphrasing

How this works

Deterministic marks can be removed exactly. Probabilistic patterns can only be reduced. Simple Unmark keeps the two separate and does not promise detector outcomes.

  • SynthID Text is applied as a logits processor during generation: Google describes it as augmenting the model's logits with a pseudorandom g-function. The watermark is therefore a property of which words were chosen, not an extra character you can find and delete.
  • That is why searching Gemini output for a hidden symbol finds nothing. There is nothing to find — and equally, cleaning invisible characters does nothing to the SynthID signal.
  • Google states that detector confidence can be greatly reduced when text is thoroughly rewritten, and that the signal is weaker on short and highly factual passages. Rewriting is the documented lever, which is what this page's tool does.

Three steps

  1. 1

    Paste the Gemini text. The browser-side scan tells you whether the copy also carries invisible characters from the page or editor you copied through.

  2. 2

    Run the clean. The deterministic pass handles the character layer; the rewrite replaces the original token-choice pattern with a new one.

  3. 3

    Read the rewrite against the original before you use it, then copy the result.

Worked example

Before

Gemini's summary paragraph, pasted straight from the chat window.

After

The same content, restructured: different clause order, transitions and vocabulary.

Nothing visible is deleted from Gemini text by the deterministic pass, because SynthID Text adds no characters. The change that matters here is the rewrite, and its output differs on every request, so no fixed sample is shown.

Scope

What this page can and cannot do

Can

  • Replace the original token-choice pattern with a fresh one, which is the mechanism Google identifies as reducing detector confidence
  • Clean any invisible characters the text picked up from the page, editor or CMS it travelled through
  • Explain the mechanism honestly, with Google's own documentation cited and dated

Cannot

  • Delete SynthID Text like metadata — there is no character or field to strip
  • Report a SynthID score, because Google does not offer a public detector for arbitrary text
  • Guarantee a result, since Google itself describes reduction in confidence rather than removal
  • Touch SynthID in images, audio or video, which are separate systems

Submitted text and cleaned output are not stored in our application database, and no text is ever sent to analytics. Deterministic inspection on the invisible character remover runs entirely in your browser. See the privacy policy and methodology.

Questions

About gemini watermark remover for text

Sources

Does Gemini watermark text?
Google documents SynthID Text as its approach to watermarking AI-generated text and has described deployment at scale in Gemini, along with an open-source reference implementation. See the dated status block on this page for what is first-party documented versus reported.
Can a SynthID text watermark be deleted like metadata?
No. A statistical token-choice pattern is part of the wording. There is no field or hidden character to strip, so meaningful reduction requires genuinely new phrasing.
Can I check whether the watermark is gone?
Not with this tool, and not with any public tool we are aware of for arbitrary Gemini text. Google publishes the detection approach but not a general consumer detector, so anyone claiming to verify SynthID removal for you should be treated sceptically.
Does this work on Gemini-generated images?
No. This is a text-only tool. SynthID also marks images, audio and video, and those require entirely different handling.

Other cleaners

Need more than three cleans?

Credits cost 0.1 per started 100 words, never expire, and there is no subscription. An account raises the limit to 5,000 words per clean.

See pricing