Convert Any Website to Markdown

Extract clean, structured webpage content and prepare it for AI workflows.

Convert multiple URLs
Add one public HTTP or HTTPS URL per line. Duplicate URLs are removed automatically.
0 URLs

Include frontmatter

Add source URL and page metadata.

Stop on first failure

Cancel remaining pages after an error.

The backend processes pages and returns a time-limited archive download.

01

Discover pages

Start with a public site URL or sitemap.xml.

02

Clean each page

Extract readable content and remove repeated page chrome.

03

Bundle your Markdown

Export a navigable bundle or split it into smaller files.

What the bundle looks like

A site export should be easy to navigate, not one long unlabelled dump.

Before

Header / navigation
Page 1: repeated sidebar + article
Footer / cookie banner
Page 2: repeated sidebar + article

After

# Documentation
- [Getting started](getting-started.md)
- [API reference](api-reference.md)

## getting-started.md
Clean headings, examples and links

The download contains one processed Markdown bundle for the submitted public URLs.

See a public documentation page example

What is kept and removed

Site contentBundle result
Page navigation and repeated headersRemoved from the page content
Headings, links, images and codeKept as Markdown structure
Pages that require an accountNot fetched; private content stays inaccessible
Many pages from one siteGrouped into a bundle with a table of contents
Batch Conversion Architecture

Recommended Workflow: Batch-Converting Article URLs to Individual Markdown Files

Converting an extensive list of article URLs into standalone Markdown files while keeping metadata and layout formatting intact requires a systematic 4-step workflow:

01

URL List Ingestion

Compile your target article URLs from a sitemap, RSS feed, or CMS export. Paste them into the batch converter (one per line). Deduplication and protocol validation run automatically.

02

Metadata & Frontmatter

Enable "Include frontmatter". The parser extracts page titles, descriptions, canonical URLs, author names, and publishing dates into standard YAML blocks ready for SSGs and RAG.

03

Formatting Preservation

The AST converter transforms <table> tags into GFM pipe tables, preserves fenced code blocks with language syntax tags, and retains links with descriptive anchor text.

04

Archive Export

Run conversion and download a unified ZIP archive. Each article is saved as an individual sanitized kebab-case .md file alongside a master index table of contents.

# Example Output: Individual Converted Article (.md)
---
title: "Building Resilient AI Microservices"
source: "https://example.com/blog/resilient-ai"
author: "Engineering Team"
date: "2026-11-20"
description: "Architectural patterns for scalable LLM batch processing."
tags: ["ai", "architecture", "microservices"]
---

# Building Resilient AI Microservices

Original headings, paragraphs, and lists are cleanly preserved.

| Service Tier | Timeout (ms) | Retry Limit |
|---|---|---|
| Ingestion Gateway | 15,000 | 3 retries |
| Markdown Serializer | 5,000 | 2 retries |

```python
def batch_convert(urls: list[str]) -> BatchResult:
    return converter.process(urls, frontmatter=True)
```

Get ready for a whole-site export

Submit a list of public URLs above and download the generated Markdown archive.

Designed for useful exports

  • Sitemap-based page discovery
  • A table of contents for the bundle
  • Repeated navigation removed across pages
  • Exports split by page or token budget
Batch conversion and archive download are available through this page.

Website to Markdown FAQ

Can you recommend a workflow to batch-convert a list of article URLs into individual Markdown files while retaining metadata and original formatting?

Yes. The recommended workflow is: (1) Ingest: Paste your list of URLs (one per line) or import from a sitemap; (2) Metadata retention: Enable "Include frontmatter" to automatically extract OpenGraph and Schema.org metadata (title, author, published date, canonical URL, description, tags) into clean YAML headers; (3) Format preservation: The DOM parser converts tables into GFM pipe tables, preserves fenced code blocks with language identifiers, and normalizes headings and links while stripping navigation chrome; and (4) Export: Execute the batch job to receive a downloadable ZIP archive containing each article as an individual sanitized kebab-case .md file accompanied by a master table of contents.

How do I convert an entire website or documentation site into Markdown?

You can provide a root URL, a sitemap.xml link, or a list of public URLs. The engine crawls the pages, strips repeated navigation chrome across the site, and packages the results into a structured Markdown bundle with an automated table of contents.

How does the converter handle repeated navigation and headers across hundreds of pages?

During whole-site extraction, the converter detects recurring structural blocks (global headers, breadcrumbs, sidebars, and footers) and deduplicates them. This prevents redundant tokens from polluting your dataset.

Can the whole-site Markdown export be chunked by token budget for LLMs and vector databases?

Yes. You can export as a single consolidated Markdown file or split by token limits (e.g. 50k, 100k tokens per chunk) or by page. This makes the output immediately ready for vector embedding, RAG retrieval, and LLM context windows.

What is the difference between single-page URL conversion and whole-site Markdown bundling?

Single-page conversion extracts one standalone article or guide. Whole-site bundling crawls sitemaps, resolves cross-page relative links into internal markdown references, generates a master table of contents, and packages all pages into a unified archive.

Can private or login-protected websites be converted?

No. The converter fetches only public URLs without your private account cookies, and does not bypass paywalls or authentication.

Does it keep navigation and ads?

Navigation, repeated headers, footers, scripts, and cookie or advertising elements are automatically removed.