Donaldson R F Extract Domains from URLs
Paste a list of links and get the domains back: the root (registrable) domain such as bbc.co.uk or example.com.br, or the full host name. Duplicates removed, sorted and counted.
Skipped inputs
How each input was read (first 300)
Why "the last two labels" is not enough
The root domain (also called the registrable domain) is the name a person or company actually registered. For www.bbc.co.uk that is bbc.co.uk, not co.uk, because co.uk is itself a suffix that anyone can register under. The same is true of com.br, com.au, co.jp and hundreds of others. There is no rule in the domain name system that tells you which endings behave like this, so browsers and many tools use the Public Suffix List, a community-maintained list of them. This tool has a copy of that list built in and applies its official matching algorithm: the longest matching rule wins, wildcard rules (such as *.ck) and exception rules (such as !www.ck) are honoured, and an ending that is not in the list is treated as unknown rather than guessed.
What is read from each input. The scheme (https://), a user name and password, the port, the path, the query and the #fragment are ignored; only the host name is used, converted to lower case. A line without a scheme, such as shop.example.co.uk/basket, is read as a host name. mailto: links and e-mail addresses give the domain after the @. International names are compared in their xn-- (punycode) form so that the same site is never counted twice. A trailing dot (example.com.) is removed.
Public and private suffixes. The list has two parts. The ICANN part holds real registry suffixes (com, co.uk). The private part holds suffixes that companies run for their customers, such as github.io or blogspot.com, where every customer is a separate site. By default the private part is ignored, so me.github.io gives github.io; tick the box above to get me.github.io instead. Choose whichever matches what you mean by "the site".
What is skipped, and why. Numeric addresses (192.168.0.1, [::1]) have no domain, single-word hosts such as localhost and endings that are not in the list (a typo like example.con, or a file name like report.pdf) are not real domains as far as the list knows, and an input that is only a public suffix (co.uk) has no registrable part. Each is listed under "Skipped inputs" so you can see what was left out; two of these can be switched on above. The list changes: new endings appear and private suffixes come and go, so a snapshot can be a little out of date. For a list of your own e-mail addresses, see Count Emails by Domain; to look a domain up afterwards use Whois Lookup or DNS Lookup, and to break one link into parts try the URL Parser.
Source / Last verified (26 September 2026): Public Suffix List (Mozilla Public License 2.0), snapshot version 2026-09-24_13-26-36_UTC, commit a179a48, with 6,950 ICANN and 3,384 private rules. The matching algorithm is the one on the list's format page, and the tool was checked against the list's own published test cases.