URL to Markdown

Turn any public web page into clean Markdown. Keep the article text, headings, links, tables and images, drop menus and ads, then copy or download .md or JSON.

Uses a server. The address you enter is sent to our server, which downloads that page and converts it. Results are cached for up to 6 hours. Cloudflare verifies requests to help prevent abuse.

Works with articles, blog posts, documentation and other public pages.

Checking your browser before converting…

Paste a web page address to begin.

Extracts the main text of one page and drops menus, ads and footers. Pages that need JavaScript or a login, private network addresses and files other than HTML or plain text are not supported.

How to convert a web page to Markdown

  1. Paste the address of a public page, such as an article, a blog post or a documentation page. You can leave out https://.
  2. Wait for the human verification check to finish, then press Convert.
  3. Turn Include images or Include links off if you only need the text. The result updates instantly, without fetching the page again.
  4. Choose Markdown, or JSON to get the title, author, date and source address together with the text.
  5. Press Copy, or Download to save a .md or .json file named after the page title.

Clean text for AI assistants, notes and documentation

Web pages are full of menus, cookie banners, ads, sidebars and scripts. The converter keeps only the main content: headings, paragraphs, lists, tables, code blocks, links and images, written as Markdown. Image and link addresses are made absolute, so they still work outside the original site.

Markdown is compact and easy for language models to read, so a converted article uses far fewer tokens than raw HTML. Paste it into ChatGPT, Claude or another assistant to summarize or translate a page, add it to a RAG knowledge base, or save it in Obsidian, Notion or a Git repository. The token estimate assumes about four characters per token; exact counts depend on the model.

What it cannot convert

The converter reads the HTML that the website sends and does not run JavaScript, so apps that build their content in the browser may come back empty. Pages behind a login or a paywall, private and intranet addresses, and files such as PDFs or images are not supported. It converts one page at a time and does not crawl a whole site.

Some websites block automated requests. When that happens, the converter tries again through a backup network; if the site still refuses, try again later.

Limits

Each conversion needs a new human verification. You can convert up to 6 pages per minute and 60 per hour. Pages larger than 5 MB of HTML, or that produce more than 1 million characters of Markdown, are rejected rather than cut short.

Where does my data go?

This tool works through our server. Your browser sends it the page address and a Cloudflare verification token, and your IP address is used to verify the request and apply rate limits. Our server downloads the page, converts it and may cache the result for up to 6 hours. The website you convert sees a request from our server, not from you.

Like the rest of the site, this page uses Google Analytics and Yandex Metrica, including Metrica session replay. Tool events report only that the tool was opened or used.

Also useful