IPv6 Subnet Calculator
Enter an IPv6 address and prefix length to get the network address (compressed and full), the last address in the range, the total address count, and the address type — all computed in your browser.
Subnet details
- Network (compressed)
- 2001:db8:abcd:12::
- Network (full)
- 2001:0db8:abcd:0012:0000:0000:0000:0000
- Last address
- 2001:db8:abcd:12:ffff:ffff:ffff:ffff
- Prefix
- /64
- Total addresses
- 18,446,744,073,709,551,616
- Address type
- Global unicast (2000::/3)
About this tool
IPv6 subnetting works the same way as IPv4 conceptually, but the 128-bit address space is far too large for 32-bit integer math, so this tool uses arbitrary-precision BigInt arithmetic to keep every result exact. Give it an address and a prefix length (for example 2001:db8::/64) and it returns the network address in both RFC 5952 compressed form and fully expanded form, the last address in the block, the total number of addresses the prefix contains, and the type of address — global unicast, unique local, link-local, multicast, and so on. Because IPv6 prefixes routinely describe astronomically large ranges (a single /64 holds over 18 quintillion addresses), the host count is shown as an exact integer rather than rounded. All parsing and math happen locally in your browser; no address is transmitted.
How to use it
- Enter an IPv6 address, using :: to compress consecutive zero groups if you like.
- Set the prefix length from 0 to 128.
- Read the network and last addresses plus the total count from the results.
- Copy all values with the copy button.
Frequently asked questions
- What does the "::" in an IPv6 address mean?
- It is shorthand for one or more consecutive groups of all-zero 16-bit blocks, and it may appear at most once in an address. For example 2001:db8::1 expands to 2001:0db8:0000:0000:0000:0000:0000:0001. This tool accepts compressed input and shows both compressed and full output.
- Why is there no broadcast or "usable host" count like IPv4?
- IPv6 has no broadcast address — it uses multicast instead — and it does not reserve the first and last addresses of a subnet the way IPv4 does. So the total address count of a prefix equals 2^(128 − prefix), with no subtraction.
- What is the canonical (RFC 5952) form?
- RFC 5952 defines a single preferred text representation: lowercase hex, leading zeros in each group removed, and the longest run of two or more zero groups compressed to "::". The "Network (compressed)" output follows this rule so addresses are directly comparable.
- What do the address types mean?
- Global unicast (2000::/3) is internet-routable; unique local (fc00::/7) is the IPv6 equivalent of private RFC 1918 space; link-local (fe80::/10) is valid only on a single link; and multicast (ff00::/8) addresses a group of interfaces. The tool infers the type from the leading bits.
- Why use a /64 prefix so often?
- A /64 is the standard subnet size in IPv6: it leaves a 64-bit interface identifier, which is what stateless address autoconfiguration (SLAAC) expects. Most LAN segments are /64 regardless of how few hosts they contain.
- Does this tool send my addresses to a server?
- No. Parsing and BigInt math run entirely in your browser. Nothing you enter is transmitted or stored, so it is safe for internal and production prefixes.