Clean Markdown. Built for models.
get-md is a fast, lightweight HTML, PDF, DOCX, and Markdown to Markdown converter optimized for LLM consumption. The standard path uses Turndown and Mozilla Readability to return clean, structured Markdown in under 100ms.
Built by the Nano Collective rather than a company, it runs as a typed library or a CLI, ships no forced telemetry, and keeps the optional AI conversion on a local model so your content never leaves your machine.
Premium capabilities.
Zero paid tiers.
Sub-100ms
Turndown and Mozilla Readability convert clean, structured Markdown in under 100ms, with no model download required.
Flexible Inputs
Pass raw HTML, PDF files, DOCX documents, Markdown files, or a URL. get-md handles fetching, retries, and content extraction for you.
Optional LLM
Enable a local ReaderLM-v2 model for higher-quality conversion when a page needs it. Entirely on your machine.
Metadata Extraction
Pulls title, author, reading time, and more into clean YAML frontmatter, ready for downstream pipelines.
Content Filtering
Toggle images, links, and tables, or strip noise entirely. Keep exactly the signal your model needs.
Library & CLI
Use it as a typed Node.js library or straight from the terminal. Same engine, whichever fits your workflow.
React Native
Full React Native support means the same converter runs in your mobile app, not just on the server.
Batch & Sitemap
Crawl an entire sitemap or a list of URLs with bounded concurrency, then write one file per page.
Realistic workflows.
get-md handles single pages, content filtering, and whole sitemaps from the terminal or your codebase.
await convertToMarkdown(
"https://example.com"
);
Install in seconds.
Zero complex setup. Available on your favorite package manager.
Why developers choose get-md
100% Open Source
Read the code. Audit the code. Improve the code. It belongs to the community, not a corporation.
Fast by Default
The standard path is pure TypeScript and ships no heavy native binaries. Convert thousands of pages without waiting.
Local LLM Option
When you do want AI-powered conversion, it runs on a local model. Your content never leaves your machine.
Built for LLMs
Output is tuned for token efficiency and clean structure, exactly what retrieval and prompting pipelines need.
Library or CLI
Drop it into a TypeScript project or script it from the shell. No lock-in to a single way of working.
Community Driven
Features are built based on what developers actually need, not what looks good in a marketing pitch.
Supported by
Sponsorship funds the people building open, privacy-respecting AI tools, and keeps every line of it free at the point of use.
Build with us.
Join developers building the future of local-first AI tooling.