About Markdownee

Markdownee fetches web pages, extracts their main content, and saves Markdown, plain text, or HTML for research, datasets, and retrieval pipelines. Collection controls choose which pages to visit; extraction controls choose what to retain.

Extraction and crawling

Trafilatura Core, our open-source extraction engine based on Trafilatura, is implemented in TypeScript and Python. Markdownee's TypeScript version uses the TypeScript engine; its Python version uses the native Python engine. Core removes boilerplate from supplied HTML; Markdownee adds fetching and output conversion.

Core's TypeScript engine ports original Python Trafilatura, with go-trafilatura as a DOM translation aid. Its native Python engine translates that port.

Crawlee drives Playwright when JavaScript-heavy pages need browser rendering before their HTML reaches Trafilatura Core. HTTP-only fetching is also available. The npm CLI/library and Apify Actor use TypeScript; the Python library uses Crawlee Python, with its supported controls documented separately.

Token-efficient output for LLMs

Removing page furniture and HTML syntax can reduce the text sent to a language model. Savings depend on the page, settings, output format, and tokenizer. Compare the extracted result with the source before indexing it; less text is useful only when it retains the material your application needs.

Use Getting started for a first extraction or Help to choose an interface. The Apify Actor provides hosted runs and scheduling. Source is on GitHub.

Who runs Markdownee?

Glueo, s.r.o., a Prague-based software company, operates Markdownee and provides custom software development.

Updated: October 3, 2026