Extract Domains from Content

Take an address like https://blog.example.co.uk/pricing/. Reading it from the right:

• The TLD (top-level domain) is the last part, .uk. Plenty of countries hand out names one level down from there, so the part a registrar actually sells you is co.uk. Our tool treats that whole ending as the "extension" and shows it in the chips above your list.
• The root domain is the name someone registered: example.co.uk. It's the ending plus one more word. This is what most people mean by "the domain", and it's what you get with the default "Root domains" setting.
• A subdomain is anything bolted onto the front, like blog., shop., or www.. Big sites spread themselves across dozens of them.
• The hostname is the full thing, blog.example.co.uk. Switch "Show" to "Full hostnames" when you need that level of detail, or pick "grouped" to see every hostname tucked under its root domain.

The reason this matters: a backlink report that says you have 400 linking domains might turn out to be 120 real websites once www., m., and a few blog subdomains are folded in. Counting at the right level changes the answer.

Most of the people who land here are doing one of a handful of jobs. SEOs paste a backlink export from Ahrefs, Semrush, or Search Console to see which sites actually link to them, how many times each one does, and which ones belong in a disavow file. The "Disavow file" format writes each one as domain:example.com, which you can hand straight to our Disavow File Generator or upload to Google as-is.

Content and PR teams paste a competitor's article, a resource page, or a whole sitemap to see who it links out to, which is a quick way to find publications and partners worth pitching. Marketers paste a column of email addresses to find out which companies are on a list, since the domain after the @ tells you that in one step. Developers and site owners paste server logs or referrer reports to see where traffic and hotlinking are coming from, with the "IP addresses" option switched on when the log is full of them.

And a lot of it is simple cleanup: you have a messy spreadsheet of URLs in every imaginable format and you just want one tidy list of domains, sorted, with no duplicates. If what you actually want is the full page addresses rather than the domains, use the Extract URLs from Text tool instead; the two are built to work together. The CSV format here is the one to pick for spreadsheet work, since it carries the count, the extension, and every hostname that was folded into each domain as separate columns.

In "Root domains" mode the tool trims every hostname back to the name that was registered, so blog.example.com, www.example.com, and shop.example.com all count as mentions of example.com. That's usually what you want when you're asking "how many different websites are in here?"

Hosted platforms are the exception. Thousands of unrelated people have sites at someone.github.io, someone.blogspot.com, someone.wordpress.com, someone.substack.com, and so on. Collapsing them all into github.io would hide exactly the detail you're after, so by default the tool treats the account name as part of the domain. The same goes for Shopify stores, Netlify and Vercel apps, Wix and Squarespace sites, and a couple of dozen other services. If you'd rather lump them together, untick "Treat hosted sites as separate domains".

The "Ignore www." box does one small thing: in hostname mode, it stops www.example.com and example.com appearing as two different entries. Root domain mode never shows www anyway.

A note on country endings: the tool knows that co.uk, com.au, co.nz, com.br, and a few hundred similar endings are sold one level down, so shop.example.co.uk becomes example.co.uk rather than the meaningless example.co. The chips above your list use that same ending, which is why you'll see ".co.uk" as its own group instead of being lumped under ".uk".

Anything with http:// or https:// in front of it is always picked up, whatever the ending. Names written without a scheme are trickier, because "example.com" and "file.txt" look identical to a computer, so the tool checks the ending against a list of real extensions: every two-letter country code and a few hundred generic ones like .com, .org, .io, .app, and .marketing. It also deliberately skips bare names that end in a common file extension (.js, .md, .sh, .zip, .mov) because in pasted text those are nearly always files, not Jersey, Moldovan, or Saint Helena websites. If a real site gets caught by that rule, just add https:// in front of it.

Other reasons a domain might be missing: an email client wrapped a long URL onto two lines, there's a space in the middle of it, or it's typed with a mistake like a double dot. Things that look like domains but aren't, such as "e.g.", "U.S.", version numbers, or prices, are filtered out because they don't end in a valid extension.

If something slipped through that shouldn't have, the "Leave out domains containing" box takes a comma-separated list and removes anything matching it, which is also the quickest way to drop your own site from a backlink export. You can click any extension chip to narrow the list to one ending, and click it again to see everything.

Everything happens in your browser. The text you paste is never sent to our server or anyone else's, which makes it safe for client backlink reports, customer email lists, and internal logs. You can paste up to about five million characters at once, and it's processed in well under a second. Close the tab and the text is gone; only your checkbox and format choices are remembered, in your own browser.

Because nothing leaves your machine, the tool never contacts the domains it finds. It doesn't check whether they're live, look up who owns them, fetch their titles, or follow redirects, and it doesn't pull the links out of a web page for you; you paste the content in yourself. If you need a live check on the pages behind the domains, run the URLs through the Redirect Chain Checker next.