What it accepts and reports
Paste one address per line and each line is judged as IPv4, IPv6 or invalid, with a normalized form and a fully expanded form alongside. On the IPv6 side it reads the :: shorthand, IPv4-mapped forms such as ::ffff:192.168.1.1 and a zone index such as fe80::1%eth0. Bracketed URL forms and a trailing port number are split off before the check and noted in the remarks column.
How normalization and type labels are derived
IPv6 normalization follows the IETF RFC 5952 recommendation: drop leading zeros in each group, write hex in lower case, and collapse only the longest run of zero groups with ::. The expanded form always prints eight four-digit groups. Type labels come only from ranges the RFCs define, with private IPv4 ranges from RFC 1918 and documentation ranges from RFC 5737 and 2001:db8::/32. Checked October 2026.
Limits and what it will not do
An octet with a leading zero is reported invalid because some implementations read it as octal. Parsers differ in what they accept, so this verdict may not match a particular library or operating system. The tool never connects to an address or resolves a domain name to see whether it is live, and it handles up to 200 lines at a time.
Frequently Asked Questions
A leading zero in an octet is read as octal by some implementations, so the same text can mean two different addresses. This tool treats that ambiguous notation as invalid.
It is an IPv6 address in the IPv4-mapped form, which carries an IPv4 address inside it. The verdict is IPv6 and the embedded IPv4 is shown in the remarks.
It only checks notation. It does not test whether the address responds, or look up which organization it is allocated to.