Skip to main content

About Markdownee

What is Markdownee?

Markdownee collects web content for text analysis, retrieval, and dataset preparation. Enter a URL in the online playground to inspect an extraction or configure a run. Crawlee fetches pages through Playwright or Cheerio; Trafilaturacore selects and cleans their content. You control link following, extraction options, and output formats.

Why Trafilaturacore?

Trafilaturacore is our open-source pure-TypeScript port of Trafilatura. Its extraction core is a direct port of Python Trafilatura, with go-trafilatura used only as a DOM translation aid, and it applies Trafilatura's own heuristics to strip page clutter. Markdownee consumes it as a dependency. The engine runs offline — HTML in, cleaned content out — with no Python runtime, no GPU, and nothing to compile.

Token-efficient output for LLMs

Removing page furniture and converting markup can reduce the text a language model receives. The saving depends on the page, extraction settings, output format, and tokenizer. Compare the result with the source before indexing or embedding it: fewer tokens are useful only if the extraction retains the material your application needs.

Configure collection and output

  • Boilerplate removal is powered by Trafilaturacore, our open-source pure-TypeScript port of Trafilatura. Its extraction core is a direct port of Python Trafilatura — with go-trafilatura used only as a DOM translation aid — and applies Trafilatura's own heuristics to strip navigation, sidebars, footers, and similar clutter
  • Use Crawlee with Playwright to render JavaScript content before extraction, or select Cheerio for HTTP-only fetching
  • Run extraction in Node.js without a Python runtime or GPU
  • Choose the npm CLI, npm library, or hosted Apify Actor; the open-source code is on GitHub
  • Download image files when the selected destination supports them
  • Enable Markdown discovery to look for a site's published Markdown representation and use it as the source for the requested formats

Put the extracted content to work

  • Assemble page text for a training-data preparation workflow.
  • Supply content to your own retrieval-augmented generation pipeline.
  • Collect articles for research and text analysis.
  • Run repeated extractions from your scheduler and compare their contents in a change-monitoring system.

Run a hosted crawl

The Markdownee Apify Actor provides hosted execution and storage. Configure browser rendering, extraction mode, starting URLs, and link-following limits for each run.

The company behind Markdownee

Markdownee is operated by Glueo, s.r.o., a Prague-based software company that runs its own online services, such as Markdownee, and provides custom software development.

Updated: September 7, 2026