Frequently Asked Questions

The two most common surprises aren't caused by the Text Cleaner itself, but by how text from an AI tool ends up on your clipboard in the first place: paragraphs getting lost when you copy (see the next question), and Markdown formatting like **asterisks**, which the Text Cleaner has removed automatically since version 2.5 (see the question after that). Both are intentional behavior, not bugs.

AI tools often format their replies with real paragraphs – in the underlying HTML, each one is its own block. If you select the text with your mouse and copy it with Ctrl+C, the browser turns those paragraph boundaries into a single line break instead of a blank line. The Text Cleaner's "Keep paragraphs" mode, however, treats a single line break as a wrap inside a paragraph (e.g. caused by word wrapping) and merges it into one continuous sentence – only a genuine blank line counts as the end of a paragraph.

Hi team,

thanks for the update.

I'll follow up on this tomorrow.

Hi team,
thanks for the update.
I'll follow up on this tomorrow.

The fix: use the copy icon directly below the AI tool's reply instead of manually selecting the text. It copies the raw text with correct paragraph separators to your clipboard.

The copy icon from the previous question gives you clean paragraphs, but also the raw Markdown source – including formatting characters like **bold**, *italic*, # headings, [links](url), `code`, ~~strikethrough~~, or > quotes, which AI tools normally only render visually without showing the characters themselves. Since version 2.5, the Text Cleaner removes these formatting characters automatically – in all three modes, regardless of whether you picked Clean text (preserve paragraphs), Clean text (create continuous text), or Clean text (preserve layout). What's left is plain text without asterisks and friends.

Clean text (preserve paragraphs): for running text like letters, emails, or articles. Line breaks inside a paragraph get merged into one continuous sentence, real paragraph boundaries (a blank line) are preserved. The default mode for most texts.

Clean text (create continuous text): when the entire text should end up as a single continuous block with no paragraphs at all – e.g. for form fields with limited space.

Clean text (preserve layout): when your text contains bullet points, numbered lists, or a line structure you want to keep exactly as-is. This mode doesn't merge any line breaks and leaves bullet characters like -, *, or 1. in place.

Yes. The Text Cleaner runs entirely in your browser – the text you enter is never sent to a server, neither for cleaning nor for storage. There's no tracking and no cookies. More details in the Privacy Policy.

Some watermarking methods leave no character-level trace in the text and therefore can't be removed by text cleaning:

• SynthID Text (Google DeepMind): influences token selection during generation via a secret key list, without changing any individual characters.
• Statistical token watermarking (KGW method, OpenAI research): favors "green" tokens during generation – statistically detectable, but invisible at the character level.
• Semantic watermarks: synonym choices or sentence-structure variants following a secret scheme, removable only by fully rewriting the text.
• File metadata (C2PA/Content Credentials): attached to the file itself (e.g. a .docx file), not the text, and isn't copied along when you copy the text.