Website to Markdown for AI & LLMs (RAG-ready)

Turn web pages into clean Markdown that ChatGPT, Claude, Gemini, RAG pipelines and AI agents can read. A real browser renders the page (JavaScript sites work), the main content is extracted with Mozilla Readability (the engine behind Firefox Reader View), and menus, ads, footers and cookie banners are dropped. Headings, tables, code blocks, lists and links stay.

• ✅ Clean main content or the whole page without navigation and footer
• ✅ GitHub-flavored Markdown: tables, fenced code, lists, absolute links
• ✅ Token estimate per page (approxTokens), word and character counts, optional length limit
• ✅ Metadata: title, description, site name, language, publish date
• ✅ Works on modern sites: JavaScript rendering, strict security policies and AMD/RequireJS pages
• ✅ Fair billing: failed pages, error pages (4xx/5xx) and pages excluded by robots.txt are free

What can you use it for?

• RAG and knowledge bases: feed documentation, help centers and articles into a vector database
• AI agents: let an agent read any page through the Apify API or MCP and get compact text instead of raw HTML
• Summaries and research: collect articles for summarization, translation or analysis
• Content migration: move pages from an old CMS into Markdown-based docs

Input example

{
    "urls": ["https://github.com/mozilla/readability", "https://en.wikipedia.org/wiki/Markdown"],
    "mode": "article",
    "includeLinks": true,
    "includeImages": false
}

Output example

Real output from a test run (Markdown shortened):

{
    "url": "https://github.com/mozilla/readability",
    "finalUrl": "https://github.com/mozilla/readability",
    "statusCode": 200,
    "title": "GitHub - mozilla/readability: A standalone version of the readability lib",
    "description": "A standalone version of the readability lib. Contribute to mozilla/readability development by creating an account on GitHub.",
    "siteName": "GitHub",
    "language": "en",
    "publishedTime": null,
    "mode": "article",
    "markdown": "A standalone version of the readability library used for [Firefox Reader View](https://support.mozilla.org/kb/firefox-reader-view-clutter-free-web-pages).\n\n## Installation\n\nReadability is available on npm: ...",
    "wordCount": 1025,
    "charCount": 7300,
    "approxTokens": 1825,
    "truncated": false,
    "convertedAt": "2026-10-20T21:19:49.560Z"
}

mode tells you what was used: article (main content) or full (whole page, used automatically when a page has no clear main text, like home pages and link lists). A summary per URL is saved as OUTPUT in the key-value store.

Pricing

A fixed price per converted page. In our tests 6 pages took about a minute with the default 2 GB memory.

Paid Apify plans pay less: 10% off on Starter, 20% on Scale and 30% on Business and higher plans.

FAQ

Does it crawl a whole website? No, every URL is one page. To convert a whole site, pass its page URLs (for example from its sitemap).

Why is a page "blocked_by_robots"? By default the Actor respects robots.txt. Turn off Respect robots.txt only for your own sites or with the owner's permission.

Is it legal? Converting public pages for reading, research and AI processing is common. Respect copyright and the site's terms when you store or republish content. You are responsible for your use of the data.

---

Impressum

Levi Griesinger, Tidelane Data
Besigheimer Straße 18
74369, Löchgau, Deutschland
E-Mail: LeviGriesinger@gmail.com