RunLocalTools

HTML Boilerplate Stripper — Extract Clean Article Text from Any Page

Paste raw or scraped HTML and pull out just the readable title and body text. Strips nav, footers, scripts, and ads automatically — free, and everything runs in your browser.

What it does

HTML Boilerplate Stripper takes messy, scraped, or copied HTML — the kind you get from “view source” on a real webpage — and pulls out just the readable content: the title and the actual paragraph text, minus the navigation bars, footers, ad slots, comments, and scripts. Output comes as plain text or lightweight Markdown, ready to paste into notes, a document, or another tool.

How it works

There are two ways to clean a page, and this tool gives you both:

  • Readability mode scores every paragraph-like block on the page by text density — length, punctuation, how much of it is inside links — and picks out the block that reads like the actual article, the same general approach browsers use for their built-in “reader view.” Sidebars, navigation, and link-heavy footers score low and get left behind automatically.
  • Tag strip mode instead keeps the whole page and simply removes whichever categories you’ve checked — navigation and footer elements, forms and form controls, and (in Markdown output) images.

Script and style content is never included in the output either way, regardless of which boxes are checked — raw JavaScript or CSS was never something you’d want mixed into extracted article text.

How to use it

  1. Paste HTML source into the box, or drop an .html file onto it.
  2. Pick Readability to auto-detect the main content, or Tag strip to keep the full page and remove specific sections yourself.
  3. Choose Plain text or Markdown as the output format — Markdown keeps headings, bold/italic, links, and (optionally) images.
  4. Adjust the checkboxes for nav/footer, forms, and images to fine-tune what gets left out.
  5. Copy the cleaned result or download it as a .txt or .md file.

FAQ

Why does Readability mode sometimes pick the wrong block on an unusual page layout?

The scoring heuristic looks for paragraph-dense text with a low proportion of link text — it works well on typical article and blog layouts, but a page with very little real prose (a product listing, for example) may not have a clear “main content” block to find. Switch to Tag strip mode for more direct, manual control on pages like that.

Will this leak <script> content into my output?

No — script and style content is excluded from the output unconditionally, regardless of how the checkboxes are set.

What happens with malformed or unclosed HTML?

Browsers (and this tool) are very forgiving of broken markup — an unclosed tag or two won’t stop the extraction from working; the parser recovers the same way a browser would when rendering the page.

Does it handle non-English text and emoji correctly?

Yes, the whole pipeline is Unicode-safe end to end.

Is anything ever uploaded?

No — parsing and cleanup both happen in your browser using the browser’s own HTML parser; nothing you paste is sent anywhere.

Once you have clean plain text, the word and character counter is a quick way to check length before pasting it somewhere with a limit. If what you’re actually starting from is a .docx or PDF rather than a webpage, that’s a different conversion path — this tool is specifically for HTML source.