Subnet calculator

Calculate a network, understand the bits, and plan your subnets. Each tool explains both the result and how to get there.

1. Understand an IPv4 subnet

An IPv4 address has 32 bits. The prefix (the number after /) tells you how many identify the network. The remaining bits identify addresses within it.

Addresses, capacity, and what each value means
Result	Value	What it means

How this result is calculated

See the binary calculation

Each group is an 8-bit octet. A mask has consecutive 1s for network bits, followed by 0s for host bits. AND keeps a bit only when both inputs are 1.

32-bit representations
Value	Binary

For example, 10 AND 0 = 0 in the last octet of a /24 mask, so 192.168.1.10 belongs to 192.168.1.0/24.

2. Find a subnet size for your hosts

Use this when you know how many addresses you need. Count every interface that needs an IP, including gateways, and allow for growth. This tool sizes conventional LAN subnets, reserving one network and one broadcast address.

3. Split into equal-size subnets

Use equal-size splitting when each segment needs the same capacity. A longer prefix borrows host bits to identify more networks. Every extra bit doubles the number of subnets and halves the addresses in each one.

How the split works

All generated subnets, in address order
Network / CIDR	Mask	First host	Last host	Broadcast	Total	Usable

4. Plan different-size subnets (VLSM)

Variable Length Subnet Masking gives each segment a size matched to its host requirement. It can save address space compared with giving every segment the largest subnet size.

How allocation works

For each request, add two reserved addresses, then round up to a power of two.
Place the largest blocks first, starting at the parent network. This keeps subsequent blocks aligned without overlaps.
Stop with an error if the combined blocks do not fit. Labels keep the original input order.

Allocations, largest first
Subnet	Network / CIDR	Requested	Usable	Host range

Remaining addresses are unallocated space, not a count of usable hosts in a future subnet. Each new LAN subnet also needs its own network and broadcast addresses.

5. Test whether an address is in a subnet

Use this when checking an access list, a firewall rule, or a host that cannot reach its gateway. The answer comes from the block's first and last address, so a /8 is tested as quickly as a /30 and no list of addresses is ever built.

6. Compare two networks

Use this when planning address space or chasing a routing conflict. Two valid CIDR blocks can only be separate, identical, or one inside the other. They cannot half-overlap, because every block begins on a boundary that is a multiple of its own size.

What this means

Both networks, and the addresses they share
Block	Range	Total addresses

7. Turn an address range into CIDR blocks

Tool 1 already goes the other way, turning a block into its range. This goes from a plain start and end address back to CIDR. Not every range is one block: a block must begin on a boundary that is a multiple of its own size, so a range starting or ending off a boundary needs several.

How the blocks are chosen

The blocks that cover this range, in address order
Block	Range	Total addresses

8. Summarize several networks

Summarizing replaces several routes with one shorter prefix. The risk is quiet: the covering block often takes in address space nobody asked for, and advertising it claims that space too. This tool always shows what the summary would pull in.

Keep learning

Subnet capacity is arithmetic, not permission to assign every address. Private, loopback, multicast, and other special-use ranges have their own rules, so the calculator reports an address type for whatever you enter.

That type comes from the IANA special-purpose registry rather than from the first octet. The entered address is matched against every reserved block and the most specific one wins, which is why 255.255.255.255 reports as a limited broadcast instead of simply reserved. An address matching no reserved block is public. The type describes the address you entered, not every address in its network.

Further reading: CIDR and prefix lengths (RFC 4632) · /31 point-to-point links (RFC 3021) · Private address space (RFC 1918) · Special-purpose registry (RFC 6890).