nectr is live: clean, always-fresh markdown for AI
Turn any web page or PDF into clean, LLM-ready markdown, behind a permanent link that stays up to date, with an API, extraction modes, and bot analytics.
Nectr has just launched. It allows you to get a clean, structured, markdown from any web page or PDF link for AI models to read. Markdown is served behind one permanent link that stays up to date automatically.
Why I built it
Feeding raw HTML to an LLM wastes tokens and drags in navigation, ads, and boilerplate. I needed this tool for my own projects and so nectr was created. It renders each page like a real browser, strips the clutter, and gives you just the content as markdown.
What you can do
- One always-fresh link: nectr re-crawls on a schedule, or on request if too much time has passed, so your link does not go stale. Reads are free.
- Pages and PDFs: both convert to clean markdown, tables included.
- Extract exactly what you need: clean article, full page, outline, tables, lists, or structured JSON.
- A simple API:
POST /conversionswith a Bearer key. See the docs. - See which AIs read your content: GPTBot, ClaudeBot, PerplexityBot, and more.
- Private by design: Swiss-hosted, encrypted at rest. Not used to train models.