IPv6 Address Calculator
Make sense of any IPv6 address or prefix. Enter an address with its prefix length to see the compressed and expanded forms, the network it belongs to, the range of addresses, the number of /64 subnets and the reverse DNS zone, and split the prefix into smaller networks.
- Runs in your browser
- No sign-up
- Free to use
Add a prefix length after a slash, for example /48. Without one, /64 is assumed.
How to use IPv6 Address Calculator
- Enter an IPv6 address, optionally followed by a slash and a prefix length such as /48.
- Read the network prefix and the address range at the top.
- Use the Address and Network sections for the expanded form, type, reverse DNS name and counts.
- Choose a prefix under Split into subnets to list the smaller networks inside.
IPv6 Address Calculator features
Expand and compress
The full eight-group form and the canonical short form recommended by RFC 5952.
Network range
The network prefix with its first and last address, compressed and expanded.
Exact large numbers
Address and subnet counts calculated exactly, with powers of two for the very large ones.
Subnet splitting
Lists the subnets of any longer prefix, with a copy button for each.
Address type
Global unicast, unique local, link-local, multicast, documentation, 6to4, Teredo and other special ranges.
EUI-64 and reverse DNS
Recovers the MAC address from an EUI-64 interface identifier and builds the ip6.arpa name.
When to use IPv6 Address Calculator
- Planning how to divide a /48 from a provider into /56 or /64 networks for sites and VLANs.
- Writing the reverse DNS zone name for a delegated prefix.
- Checking whether two addresses fall into the same network.
- Normalising addresses to the canonical short form before comparing or storing them.
IPv6 Address Calculator FAQ
How is an IPv6 address shortened?
Two rules apply. Leading zeros in each group may be dropped, so 0db8 becomes db8. One run of consecutive all-zero groups may be replaced by a double colon. To keep the address unambiguous the double colon can appear only once; the canonical form uses it for the longest run and writes letters in lower case.
Why is /64 the standard subnet size?
Automatic address configuration (SLAAC) and several other IPv6 features expect the last 64 bits to be the interface identifier. A /64 holds about 18 quintillion addresses, far more than any network needs, but size is not the point: the fixed boundary keeps every network uniform. Use /64 for every ordinary LAN.
How many /64 networks are in a /48 or a /56?
A /48 contains 65,536 networks of size /64, and a /56 contains 256. Each bit you move the prefix adds a factor of two, and each hexadecimal digit, four bits, a factor of sixteen.
What are the address types?
Global unicast addresses (2000::/3) are public and routable. Unique local addresses (fc00::/7, in practice fd00::/8) are the private range. Link-local addresses (fe80::/10) exist on every interface and work only within one network segment. Multicast addresses start with ff. 2001:db8::/32 is reserved for documentation and examples.
What is an EUI-64 interface identifier?
A way of building the last 64 bits of an address from the network card's MAC address, by inserting ff:fe in the middle and flipping one bit. When an address has this pattern, the calculator shows the original MAC address. Modern systems usually prefer random identifiers for privacy.
Why split on 4-bit boundaries?
Each hexadecimal digit represents four bits. Prefixes such as /48, /52, /56 and /60 end exactly between two digits, which makes addresses readable and lets reverse DNS zones be delegated cleanly. Other prefix lengths are valid but harder to work with.
Planning with IPv6 prefixes
IPv6 addresses are 128 bits long and written as eight groups of four hexadecimal digits. The length changes how address planning works. In IPv4, the scarce resource is addresses, and subnets are sized to fit the number of hosts. In IPv6, a single LAN already has more addresses than could ever be used, so the planning unit becomes the subnet itself, and the question is how many networks you can create and how to number them sensibly.
A typical allocation illustrates this. A provider assigns a business a /48. The first 48 bits are fixed, the last 64 bits belong to the hosts, and the 16 bits in between are yours to number 65,536 networks. Many organisations use the first hexadecimal digit of those 16 bits for the site, the next for the type of network, and the rest for the VLAN number, so that an address can be read at a glance. Homes often receive a /56, enough for 256 networks.
Because the numbers are too large for intuition, a calculator matters more than in IPv4. It shows where a prefix begins and ends, which matters when writing firewall rules and route filters, and it lists the subnets of a split so that you can assign them without overlap. Sticking to four-bit boundaries keeps every subnet aligned to a hexadecimal digit, and the reverse DNS zone shown here can then be delegated without awkward workarounds.
One habit carries over from IPv4 and should be dropped: saving addresses. Assigning a /64 to a point-to-point link or to a network with three devices is normal and intended. The exceptions are /127 for router links and /128 for loopback addresses.