6 GitHub repos to turn any website into clean, structured data:

1️⃣ firecrawl (190k+ stars)

Crawls a site and hands your AI agent clean markdown or structured JSON. The default answer in 2026.

https://github.com/firecrawl/firecrawl

2️⃣ markitdown (189k+ stars)

Microsoft's converter: PDF, Word, Excel, PowerPoint, images and HTML into Markdown an LLM can read.

https://github.com/microsoft/markitdown

3️⃣ crawl4ai (85k+ stars)

Open-source crawler built for LLM pipelines. Any website into LLM-ready Markdown, self-hosted.

https://github.com/unclecode/crawl4ai

4️⃣ docling (68k+ stars)

Parses complex documents (tables, layouts, scanned pages) and gets them ready for gen AI.

https://github.com/docling-project/docling

5️⃣ scrapegraph-ai (31k+ stars)

You describe what you want in a prompt; it builds the scraping pipeline as a graph and runs it.

https://github.com/ScrapeGraphAI/Scrapegraph-ai

6️⃣ maxun (17k+ stars)

No-code: train a robot on a website in clicks, get a structured API or spreadsheet out.

https://github.com/getmaxun/maxun

------------------------------------------------------------------------------------------------

Discover the best GitHub repos without the endless scrolling.

Get top repos organized by category (AI, DevOps, Databases & more), plus key takeaways from developers' most useful comments - all in one free, twice-weekly newsletter.

👉 Get the best repos & community insights - Free