claudewatermarks.pro

Detect and rewrite AI-writing patterns

Paste text generated by Claude to see which patterns make it look AI-written, then rewrite it to read more naturally. Free to try, no signup required.

Try the detectorTry the removerClean invisible characters (free)

Feature 1

Watermark Detector

Checks a piece of text for the writing patterns that make it look AI-generated, and shows you exactly what's driving that read. Not a guess, a breakdown.

  1. 1Paste text you suspect is AI-written (up to 500 words free, 1,000 words signed in).
  2. 2We score it using writing-pattern heuristics (sentence rhythm, stock phrasing, structure) plus an AI pass that reads the text for context-specific tells.
  3. 3You get an overall score, the text with suspect passages highlighted, and a plain-language reason for each flag.
Try the detector

Feature 2

Watermark Remover

Takes text that reads as AI-written and rewrites it to sound more natural, targeting the specific patterns the detector would flag, without changing what it says.

  1. 1Paste the text you want to rewrite (same word limits as the detector).
  2. 2We rewrite it to reduce the patterns that make text read as AI-generated: varying sentence rhythm, swapping stock phrases, breaking up uniform structure. Your meaning and tone stay intact.
  3. 3You get clean, editable output back: copy it, download it as a .txt file, or keep tweaking it right in the browser.
Try the remover

Feature 3: free, no limit

Invisible Character Remover

Finds and strips invisible Unicode characters (zero-width spaces and hidden tag data) that have no legitimate use in typed text. Runs instantly in your browser, with no signup, no server call, and no usage limit at all.

  1. 1Paste any text. Nothing is sent to a server, so there's no word limit and no usage cost, ever.
  2. 2We scan for zero-width characters and hidden Unicode tag data that never appear in normal typed text.
  3. 3You get a breakdown of exactly what was found, plus a cleaned copy with them stripped out.
Try it free

Anthropic added a real invisible watermark to Claude’s output in August 2026, but their official detector is in private preview, limited to regulators, researchers, and media, not available to the public. This tool analyzes general AI-writing patterns as a practical stand-in, not Anthropic’s cryptographic watermark.