Free Subnet Calculator (+ Guide)

What is a subnet calculator?

A subnet calculator turns an IP address and a prefix into the block it belongs to: where it starts and ends, and how many devices it can number.

This one works for IPv4 and IPv6, splits a block into smaller ones and finds the prefix for a number of hosts.

How it works

Everything is worked out in your browser, from the address you type.

Prefixes and masks

RFC 4632 replaced the old network classes with prefixes. In 172.16.0.0/16, the /16 means the first 16 bits are the network and the other 16 number the hosts, the same as the mask 255.255.0.0.

A prefix can describe any power-of-two block, from 1 address to all 2^32 IPv4 addresses. These are the common ones.

The usable hosts are the addresses minus the network and the broadcast. As RFC 3021 puts it, a /30 "requires four addresses per link": two hosts, one all-zeros network and one all-ones broadcast.

/31 and /32

A /31 has only 2 addresses. RFC 3021 lets point-to-point links use them both as hosts, since a link between two routers has two ends and no broadcast.

A /32 is one address, used for a single host route or a loopback.

Private and special addresses

RFC 1918 sets aside three private ranges for networks that don't route on the internet, and IANA's special-purpose registry lists the rest.

• Private: 10.0.0.0/8, 172.16.0.0/12 and 192.168.0.0/16
• Shared address space: 100.64.0.0/10, for carrier-grade NAT (RFC 6598)
• Loopback: 127.0.0.0/8, the machine itself
• Link local: 169.254.0.0/16, self-assigned when no server answers (RFC 3927)
• Documentation: 192.0.2.0/24, 198.51.100.0/24 and 203.0.113.0/24, for examples like the ones on this page (RFC 5737)
• Multicast: 224.0.0.0 to 239.255.255.255 (RFC 5771)

The calculator names the range for every address you type.

IPv6 subnets

An IPv6 address is 128 bits, written as 8 groups of 4 hex digits, RFC 4291 says. Unicast interface IDs are 64 bits long, so a subnet is a /64, and a /48 holds 65,536 of them.

The calculator writes addresses the way RFC 5952 asks: lowercase, no leading zeros, and :: for the longest run of zero groups.

A worked example

Here's the example the calculator loads: 192.168.10.130/26.

A /26 keeps 26 network bits and leaves 6 host bits, so the block has 64 addresses.

130 falls in the block from 192.168.10.128 to 192.168.10.191. The network is .128, the broadcast .191, and the 62 hosts run from .129 to .190.

Split into /28 blocks, it makes 4 subnets of 16 addresses and 14 hosts each. For 50 hosts, a /26 is the smallest that fits, with 62.

Mistakes to avoid

These are the slips that come up most.

• Counting every address as a host, when the network and broadcast are taken on a normal subnet
• Typing a mask with gaps, such as 255.0.255.0, which isn't a valid prefix
• Using a documentation range, such as 192.0.2.0/24, on a real network

Who is the subnet calculator for?

It's for network admins, students studying for networking exams and anyone setting up a home or office network, and these tools help around it.

• Domain checker – for a domain's DNS and host
• Scientific calculator – for powers of two
• Hex to RGB converter – for hex in another job

Draw the bits once, and the subnet answers itself.