Sitemap to LLM Markdown & RAG Pipeline
Pricing
Pay per event
Sitemap to LLM Markdown & RAG Pipeline
Extract XML sitemaps, strip boilerplate, and convert web pages into clean, token-counted Markdown for LLM ingestion and vector databases under 128 MB RAM.
Pricing
Pay per event
Rating
0.0
(0)
Developer
Piotr Hładik
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
20 days ago
Last modified
Categories
Security
Limited permissionsShare
Sitemap to LLM Markdown is an ultra-fast, async Python Actor engineered for AI developers, RAG pipelines, and vector knowledge base ingestion. It crawls XML sitemaps, fetches web pages concurrently, strips ads and navigation boilerplate with a high-performance C-based HTML parser, and outputs clean, token-counted Markdown documents.
⚡ Key Features
- Sitemap Index & Gzip Support: Seamlessly processes standard
sitemap.xml, multi-tiersitemap_index.xml, and gzip-compressed.xml.gzfeeds. - Aggressive Boilerplate Cleaning: Automatically detects and strips headers, footers, navigation trees, cookie banners, script tags, SVG icons, and advertisements.
- LLM-Ready Formatting: Converts HTML semantic tags (
h1-h6,p,pre/code,ul/ol,table,blockquote) into clean GitHub Flavored Markdown. - Accurate Token Estimation: Estimates LLM token counts and character length for every page, allowing vector embedding chunking on ingestion.
- Pay-Per-Event (PPE) Monetization: Pay only for successful page conversions ($0.001 / page) with zero idle compute waste.
- Ultra-Low Memory Footprint: Runs strictly under 128 MB RAM using async HTTP/2 connections and C-level parsing (
selectolax.lexbor).
📥 Input Parameters
| Parameter | Type | Default | Description |
|---|---|---|---|
sitemap_urls | Array | ["https://apify.com/sitemap.xml"] | Target sitemap URLs to crawl. |
max_pages | Integer | 10 | Maximum number of pages to convert. |
include_url_regex | String | None | Optional regex pattern to filter included URLs. |
exclude_url_regex | String | None | Optional regex pattern to exclude certain URLs. |
concurrency | Integer | 5 | Concurrent HTTP/2 download connections. |
proxy_configuration | Object | None | Optional proxy settings. |
📤 Output Format
Each record in the default dataset conforms to the following schema:
{"url": "https://example.com/docs/getting-started","title": "Getting Started with Python API","description": "Learn how to integrate the API in 5 minutes.","markdown": "# Getting Started with Python API\n\nLearn how to integrate the API in 5 minutes.\n\n## Installation\n\n```bash\npip install mypackage\n```","character_count": 1420,"estimated_tokens": 355,"extracted_at": "2026-10-18T16:10:00Z"}
💰 Pricing & Monetization
- Startup Event (
apify-actor-start): $0.00005 - Per Page Event (
page-converted): $0.0010