Tags out. The words left standing.
Free Claude strip HTML tool. Removes every tag plus whole script and style blocks and decodes four entities, in your browser, with nothing uploaded.
The tool
Paste markup and this Claude strip HTML tool removes every tag and gives you back the text that was between them. Script and style blocks are deleted whole, contents included, so you do not end up reading a page of CSS where the article should be.
It runs in your browser. Nothing is uploaded, there is no length cap and no run limit. Paste a whole page if you want to.
Exactly what happens
Four different things happen to four different parts of a document, and knowing which is which saves you a confused minute afterwards.
How it works
No options to set, because this route has none.
The common surprise
The line breaks that were already in your source are kept. A document with each tag on its own line comes back with the text on separate lines, because the newline characters between the tags were never touched.
What does not happen is conversion. A paragraph tag does not become a blank line, and a list item does not become a new line. So a minified document, all on one line, comes back as one long line of text. If you need paragraph breaks back, add them yourself, or run the text through the markdown formatter afterwards to tidy the spacing.
The other surprise
Entities are decoded after the tags are removed, and that order has one visible consequence. If your source contained escaped markup, written with < and > so it would display as code rather than run as HTML, those entities become real angle brackets in the output and stay there.
That is usually what you want: escaped markup was content, not structure, and it was meant to be read. If it is not what you want, run the result through a second time and the now-real tags will be stripped.
The cause
Artifacts are the main path. An Artifact is a real HTML document, and copying from it gives you markup rather than prose, complete with attributes, classes and often a style block.
The second path is a code block. Ask for a component, a template or an email and you get markup, and the moment you want only the copy out of it you are here. The third is a paste out of a rendered page somewhere else entirely, which travels with more tags than anyone expects.
Why it matters
In a CMS field, a stray tag either renders as broken layout or is escaped and displayed as visible angle brackets to your readers. In a word count, tags and attributes count as words, so a 400-word article reports as 700.
In an email, unclosed markup pasted into a rich-text composer can swallow the rest of the message. In a plain-text field, it is simply noise that a reader has to look past.
Who it's for
Writers and editors pulling the copy out of an Artifact or a template so it can be read, counted or edited as text. SEO and content teams extracting body text from a page to check length, keyword placement or reading level.
Developers use it as a quick sanity check on what a template actually says once the structure is taken away, which is a surprisingly effective way to spot placeholder text that was never replaced.
FAQ
Everything below is answered in full, nothing to click open.
Paste the markup in and run it. Every tag is removed and the text between the tags comes back.
All of them. Opening and closing tags, with their attributes, whatever the tag is called.
Line breaks already in the source survive. Paragraph tags are not converted into blank lines, so a minified document returns as one line.
They are deleted whole, including their contents, so no JavaScript or CSS appears in your text.
Four are: , &, < and >. Everything else, including " and the numeric forms, comes through as literal text.
Because that entity is not one of the four this tool decodes. Replace it yourself, or use a dedicated entity decoder.
Your source had escaped markup written with < and >. Those decode into real brackets after the tags are stripped. Run it again to remove them.
Yes. No run limit and no length cap, because the work happens on your own device.
No. It is processed in your browser and no request leaves the page. Your network panel will confirm it.
Yes. Note that text inside head elements, such as a title, is left behind once its tags are removed, so check the top of the result.
The link text stays, the href does not. The tag is removed with its attributes.
No. Alt text lives in an attribute, and attributes go with the tag.
A normal comment is removed with the rest. A comment containing a stray angle bracket can leave a fragment behind.
No. This route has no options. It removes every tag.
No. Paste text in, copy text out.
No. Markdown is a separate syntax and has its own route, the markdown stripper.
Yes. It works on markup, so whatever produced it makes no difference.
No. It removes tags rather than parsing a document, so broken markup is stripped, not repaired.
No. One document at a time, on this page.
Yes, the same way in any modern mobile browser.
Runs in your browser with nothing uploaded, no signup and no run limit.
Strip the HTML