Know What's Real in
AI-Generated Text
Wmark.ai builds focused, reliable tools for AI text detection, starting with Claude's watermark. Scan for it, or clear it, in seconds.
Try Our Tools
Two focused tools, each built for one job: detect a Claude watermark, or remove one.
Claude Watermark Detector
Paste text below to scan for Claude's invisible watermark and get a confidence score for whether it's present.
Try the Claude Watermark Detector →Claude Watermark Remover
Paste watermarked text below to clear Claude's hidden watermark. The wording is rewritten just enough to break the pattern, while the meaning stays the same.
Try the Claude Watermark Remover →Built to Be Trusted
Accurate
Built specifically to recognize Claude's watermark signature.
Fast
Results appear in seconds, with no queue and no waiting.
Free to Use
Both tools are free, with no account or subscription required.
Privacy-First
Your text is analyzed on demand and never used to train other tools.
Works Wherever You Write
Copy text from Google Docs, Notion, your CMS, or anywhere else, and paste it straight into either tool. No plugin or install required.
Build With Our API
Bringing watermark detection into your own product or workflow? API access is coming soon, get in touch to register early interest.
Contact us about API access →Frequently Asked Questions
Watermarking is an experimental technique some AI developers, including Anthropic, have researched as a way to encode a provenance signal into generated output. Our tools target the pattern associated with Claude-generated text specifically.
Detection is probabilistic, not absolute. Reliability varies with text length and how much editing has occurred since generation, so we report a confidence score rather than a flat yes/no.
No. Text you submit is processed to produce your result and is not retained longer than necessary, and it is never used to train models.
No — both tools are free to use with no account or subscription required.
The Detector evaluates text and reports whether Claude's watermark signal is present, without changing anything. The Remover acts on that result, rewriting the text to clear the signal.