How it works
Four parts, used separately
Scan is open now. Integrate, Publish and Monitor are coming.
Scan
Any public page, one at a time, without an account.
Enter a page's URL. You get back:
- The page without JavaScript: the text in its HTML
- The page as markdown
- What only appears with JavaScript, and its share of the page's text
- What the site's robots.txt says to the main AI crawlers, such as GPTBot, ClaudeBot and PerplexityBot
- Whether up to three PDFs the page links to on the same site have text, or are only scanned images
Integrate
ComingFor domains you have verified as yours.
nectr serves a clean markdown copy of each page from your own domain. Visitors and search engines get your site as it is.
- Pages and PDFs, updated when they change
- An llms.txt generated from your pages
Publish
ComingFor domains you have verified as yours.
Share a page's content, or the data on it such as a table or a list, as a dataset from your own domain. It's updated when the page changes, and you choose what's shared.
- CSV or JSON, with the source page and the date
- Public, or available by key
Monitor
ComingAnalytics and alerts.
Analytics show which AI crawlers request the copies nectr serves, and how often. They need Integrate, because nectr only sees requests for its own copies. Visits to your regular pages stay in your server's logs.
Alerts rescan your pages on a schedule and email you when one loses text. They need a verified domain.
What nectr does and doesn't do
- Scans respect robots.txt. nectr's crawler is Nectrbot.
- nectr keeps neither the page nor the result of a scan: it goes back to you only.
- nectr doesn't train AI models or sell content.
- Serving and publishing will only be for sites whose owner has verified them.
- nectr runs in Switzerland.
More in the FAQ.