Website & URL to Markdown for LLMs (HTML to Markdown)

Turn any web page into clean Markdown that's ready to feed into ChatGPT, Claude, embeddings or a RAG pipeline. Paste URLs, and get back just the article, with the navigation, ads, cookie banners and footers stripped out.

What you get, per page

• Main-content extraction: finds the article and drops menus, sidebars, share buttons, related-posts and other boilerplate.
• Clean Markdown: headings, paragraphs, lists, tables, blockquotes, code blocks, and links (made absolute) preserved.
• Metadata: title, description, author, published date, site name and language.
• RAG chunks: optional splitting on paragraph boundaries, with overlap, ready for embedding.
• A .md file per page, plus a structured JSON row for your pipeline.

Why this and not a raw HTML scraper

LLMs choke on raw HTML: it's bloated with tags, scripts and navigation that waste tokens and confuse the model. This Actor returns only the readable content as Markdown, so your prompts are smaller, cheaper and more accurate.

Who it's for

• AI and RAG builders: ingest docs, articles and knowledge bases into a vector database.
• Automation workflows (Make, Zapier, n8n, LangChain): turn a URL into clean text in one call.
• Researchers and writers: pull a readable copy of an article for summarizing or analysis.

Input

Field	Example
Page URLs	https://en.wikipedia.org/wiki/Markdown
Main content only	on
RAG chunk size	2000 characters (0 = off)

Example output (excerpt)

From the Wikipedia article on Markdown:

**Markdown** is a [lightweight markup language](https://en.wikipedia.org/wiki/Lightweight_markup_language)
for creating [formatted text](https://en.wikipedia.org/wiki/Formatted_text) using a plain-text editor.
[John Gruber](https://en.wikipedia.org/wiki/John_Gruber) created Markdown in 2004...

Each row also includes title, description, author, siteName, wordCount, optional chunks and a markdownFile key.

Pricing

Pay per page converted. Failed fetches and non-HTML URLs are never charged. The Actor respects Apify's maximum-charge-per-run setting.

Good to know

• It's a fast HTTP fetch with no browser, so it reads the HTML the server returns. Pages that render their content only with JavaScript may come back thin; for those, pair it with a browser-based crawler.
• It respects robots.txt by default and identifies itself as WebToMarkdownBot.
• For PDF files, use PDF to Text & Markdown Extractor.
• Only convert pages you have the right to use.