Gemini zero-width space remover

Gemini's answer format relies on bolded lead-ins and nested sub-points, and preserving that indentation visibly rather than letting it collapse is precisely the role a non-breaking space plays. Export the same answer to Docs and it is rebuilt a second time as an actual Google Docs document, with its own representation of spacing and structure layered on top of the first.

This tool detects and removes the zero-width space along with the related family of invisible codepoints, regardless of which rendering pass introduced them, without altering any character that actually renders. Detection and removal both run client-side, in JavaScript, with no upload and therefore no run limit.

There's nothing to preview beforehand, since the whole category of problem is invisible until the scan runs.

Step 1

Paste the reply from the chat window, or the text copied out of an exported Doc, into the input field.

Step 2

Run the removal. It executes instantly, locally.

Step 3

Read the result. Every rendered character is unchanged.

The chat window renders Gemini's nested, bold-led structure directly, and that rendering step is where a non-breaking space commonly substitutes for an ordinary one to keep visible indentation from collapsing. Export to Docs performs an entirely separate rendering pass afterward, converting the answer into a Google Docs document with its own conventions for representing the same structure.

Text copied out of the resulting Doc has therefore passed through two independent rendering steps rather than one, and either step, or both, can introduce its own invisible characters. This is a meaningfully different failure mode from a single-pass chat interface, where at least the source of any invisible character is a single rendering step.

This tool removes the zero-width space, U+200B, along with the zero-width joiner and non-joiner, the word joiner, the byte-order mark, the soft hyphen, and the non-breaking, narrow and hair spaces, the fuller family that travels with copied AI text under a name that only mentions one of its members.

A joiner detected inside a genuine emoji sequence is preserved rather than removed, since stripping it would visually split the composed emoji apart. The space-like characters in the set are replaced with an ordinary space rather than deleted outright, so word boundaries and line wrapping remain correct.

Inside a heading or a URL slug, a hidden character splits the string for any process comparing it byte by byte: full-text search, a regular expression, a string equality check. Two values that render identically fail to match, because one carries a codepoint the other does not.

Text pulled from a Docs export is a common source of this failure specifically, since a non-breaking space embedded in what appears to be ordinary indentation can silently defeat a search or a find-and-replace operating on the literal text.

These characters are overwhelmingly incidental: byproducts of nested-list rendering and the Docs export pipeline rather than markers deliberately embedded to identify a passage later. Their presence indicates that text passed through one or both rendering steps; it does not indicate who wrote the underlying words.

A genuinely different research area, statistical watermarking, biases which tokens a model is more likely to select across an entire passage, leaving a pattern distributed through ordinary word choice rather than a discrete character. No character-level scanner, including this one, can detect or affect that kind of watermark, and no tool on this site claims otherwise.

Publishing and SEO teams use it when a heading pulled from Gemini stops matching its slug, or an on-page search returns nothing for text that appears identical to what a reader sees. Anyone who exported a Gemini answer to Docs and reused it elsewhere runs the result through here to clear characters introduced by either rendering pass.

Engineers pulling Gemini output into a config file or a commit use it as well, since a hidden character embedded in an identifier produces a failure with no visible trace of its cause.