Launched this week
Paste any text and see every trace a chat interface left in it: hidden HTML class names, zero-width characters, exotic spaces, typography. Each finding has a count and a position, because these are facts about the bytes, not a probability score. Clean it in one click. Free, unlimited, no account, runs in your browser so nothing is uploaded. It does not claim to detect Anthropic's statistical watermark, because nobody outside Anthropic can. The engine is MIT open source.
Hey Product Hunt 👋 I built this after the Claude watermark news in August, when the internet filled up with tools promising to remove a watermark nobody outside Anthropic can actually detect. This does the part that is real. Copying out of a chat interface carries things with it: HTML class names, zero-width characters, exotic spaces, typography. Paste anything and you get every one of them with a count and a position, because those are facts about the bytes rather than a probability score. One click strips them. Two things I would rather say up front than bury. First, it does NOT detect Anthropic's statistical watermark. Nobody outside Anthropic can, because verification needs a key that has not been released. Any tool claiming otherwise is guessing. Second, the em dash is not the tell people think it is. I ran ten pre-computer novels through the checker: Melville uses 26 per thousand words in Moby Dick, while Austen and Stoker use none at all. A signal that swings from 0 to 26 across human authors cannot accuse anyone of anything. The detection engine is MIT open source, so you can check both of those claims instead of trusting me. Free and unlimited to check, runs in your browser, nothing uploaded. I would genuinely like to hear where it gets things wrong.
@ofir_smolinsky
Misleading. In your intro you make the honest statement that nobody can detect what Anthropic is doing yet. Since you know this, using "Claude Watermark Remover" is, at best misleading and will give newbie people who don't read this chat, will just click to your site and see the words "every trace" which is also misleading and false. You do the work to build something then damage your creditability with this clickbait-type prose. Not cool.
@rick_segal1 fair, every trace is probably too broad and im gonna tighten it. the name is based on what people search for, and i tried to be super clear on the site about what it can and cant remove, but i get how the name can still come across as misleading.
@ofir_smolinsky @rick_segal1 Great catch, Rick!
You ran ten pre-computer novels through the checker to test the em dash theory. Moby Dick hits twenty-six per thousand words, Austen none. Congrats on shipping something that leads with what it can't do.
@lucasjpols appreciate that. honestly the em dash rabbit hole was one of the reasons i shipped it. people were treating typography like proof of AI, while old books show how weak that claim is. i’d rather lead with the limitation than sell fake certainty.
You can't use "Claude" in the name, ban incoming.
Veltrix AI
I don't see any issue with cloud watermarking, but anyway, good niche findings! Good luck!
@maksym_shcherbakov1 yea i dont really have an issue with claude watermarking either lol. my issue is more with tools claiming “100% removed” when they literally cant verify that :) this just shows whats actually there, cleans it, and lets u rewrite the wording too if u wanna go further. appreciate u checking it out
minimalist phone: reduce your screentime
This is such a great response to their updates :D
@busmark_w_nika thank you :) the update created a lot of panic and a lot of tools claiming certainty they don’t have. i wanted to build the verifiable version: show what’s actually in the text, clean it, and be honest about what nobody outside Anthropic can currently prove.