Subnet Calculator: IPv4, IPv6 & CIDR Range Workstation

How to Use the Subnet Calculator

1. Select the IP version. Choose IPv4 (32 bits) or IPv6 (128 bits). The calculator remembers your last inputs for each version when you switch.
2. Enter an address. Use a bare address such as 192.168.1.10 or 2001:db8:abcd:12::1. Do not include a port, a /prefix, brackets, or a zone ID in this field. IPv4 octets must be 0–255 without leading zeros. IPv6 accepts full, compressed (::), and IPv4-mapped forms.
3. Enter a prefix or netmask.
   • IPv4: a CIDR prefix from 0 to 32 (24 or /24), or a contiguous dotted mask such as 255.255.255.0. Non-contiguous masks like 255.0.255.0 are rejected.
   • IPv6: a prefix from 0 to 128. Dotted masks do not exist in IPv6 and are rejected.
4. Read the results. The headline value is the network CIDR. The grid below it lists every derived value.
5. Inspect the bit diagram. The visualizer shows prefix bits versus host bits, with the full binary address grouped by octet (IPv4) or hextet (IPv6). Use the zoom and pan controls for 128-bit addresses.
6. Copy or print. Copy a structured plaintext summary to the clipboard, or print a clean technical cut-sheet (or save it as PDF).

Presets are included for the three most common cases: IPv4 LAN /24 (192.168.1.10/24), IPv4 point-to-point /31 (192.0.2.5/31), and IPv6 example /64 (2001:db8:abcd:12::1/64).

What Each Result Means

Subnetting Fundamentals: Network Bits and Host Bits

An IPv4 address is a 32-bit integer, normally written as four 8-bit octets. An IPv6 address is a 128-bit integer, written as eight 16-bit hextets. A prefix length $p$ splits the address into two parts:

• the first $p$ bits identify the network (the prefix, shared by every address in the subnet);
• the remaining $h$ bits identify the host or interface within that network.

The number of addresses in the block is:

This is why a subnet bit calculator reduces to powers of two: every host bit you add doubles the block, and every bit you borrow for the network halves it. A /24 has $h = 8$ and therefore $2^{8} = 256$ addresses. A /23 has $h = 9$ and $2^{9} = 512$.

Calculating the Network Address, Broadcast, and Range

Given an address $A$ and a mask $M$ (a 32-bit value with $p$ leading ones), all boundary values come from bitwise operations:

These formulas hold for IPv4 prefixes /0 through /30. The same AND operation finds the network of an IPv6 address; IPv6 simply has no broadcast address.

Netmask Calculator: Converting Between Subnet Masks and CIDR

A valid IPv4 netmask is a run of contiguous 1-bits followed by contiguous 0-bits. To convert a dotted mask to a CIDR prefix, count the 1-bits.

Example: 255.255.255.192

255      255      255      192
11111111.11111111.11111111.11000000
 8 + 8 + 8 + 2 = 26   →   /26

Going the other way, a prefix of /20 means 20 leading ones: 11111111.11111111.11110000.00000000, which is 255.255.240.0.

Only nine octet values can appear in a valid mask:

A mask like 255.0.255.0 is non-contiguous. Classless addressing (RFC 4632) requires contiguous masks, and the calculator rejects anything else.

IPv4 CIDR Cheat Sheet (/8 to /32)

CIDR (Classless Inter-Domain Routing, defined in RFC 4632) replaced the old Class A/B/C scheme. A prefix can be any length from /0 to /32. The table lists total addresses and conventional usable hosts for each prefix.

/0 (0.0.0.0/0) covers the entire IPv4 space of $2^{32} = 4{,}294{,}967{,}296$ addresses and is the default route.

Why Usable Hosts Are $2^h - 2$, and Where /31 and /32 Differ

In a conventional IPv4 subnet from /0 to /30, two addresses are reserved by convention:

• the network address (all host bits 0) identifies the subnet itself;
• the directed broadcast address (all host bits 1) reaches every host on the subnet.

Two prefixes break this rule, and a good ip subnet calculator must handle them correctly:

• /31 (RFC 3021). A /31 has only two addresses. Under the old rule it would have zero usable hosts, but RFC 3021 defines /31 for point-to-point links: both addresses are valid endpoints and there is no broadcast address. This saves half the addresses compared to a /30 on router-to-router links. For 192.0.2.5/31, the pair is 192.0.2.4 and 192.0.2.5, both usable.
• /32. A single address. It is used for loopbacks, host routes, and firewall or ACL entries that match exactly one device. First and last address are identical, and there is no broadcast.

Many older devices still assume /30 for point-to-point links, so confirm /31 support on both ends before deploying it.

Worked Examples

Example 1: A Typical LAN, 192.168.1.10/24

The first 24 bits (11000000.10101000.00000001) are the network portion. The last 8 bits (00001010, decimal 10) are the host portion.

Example 2: A Non-Octet Boundary, 10.20.37.200/21

When the prefix does not fall on an octet boundary, use the block size method. The interesting octet is the one where the mask is neither 0 nor 255. For /21 the mask is 255.255.248.0, so the third octet is interesting.

Subnets in the third octet start at multiples of 8: 0, 8, 16, 24, 32, 40, ... The address has third octet 37, which lies in the block starting at 32.

Example 3: A Point-to-Point Link, 192.0.2.5/31

Example 4: A Small Subnet, 172.16.5.130/26

Mask 255.255.255.192, block size $256 - 192 = 64$. Fourth-octet blocks start at 0, 64, 128, 192. Since 130 falls in the block starting at 128:

Example 5: An IPv6 /64, 2001:db8:abcd:12::1/64

How to Find an IP Range from a Subnet

To get the ip range from subnet notation, follow these steps:

1. Write the mask as /p and compute $h = 32 - p$.
2. Find the network address by zeroing the last $h$ bits (or by using the block-size method above).
3. Add $2^{h} - 1$ to the network address to get the last address (the broadcast for /0–/30).
4. For conventional subnets, the usable range is network $+,1$ through last $-,1$.

For /31, the range is both addresses. For /32, the range is the single address. For IPv6, the block runs from the network address to network $+,2^{h} - 1$ and is described by the prefix itself, not by "usable host" arithmetic.

IPv6 Subnetting

The same prefix logic applies to IPv6 with $N = 128$, but the practice differs in important ways (RFC 4291):

• No broadcast. IPv6 uses multicast and anycast instead. There is no directed broadcast address, which is why the calculator reports "Not applicable".
• No "minus 2" host rule. Boundary addresses are mathematical. The all-zeros address in a subnet is reserved as the Subnet-Router anycast address, but IPv6 does not define a usable-host range by subtracting two.
• The /64 boundary. For global unicast addresses, the interface identifier is 64 bits, so LAN subnets are /64. Stateless address autoconfiguration (SLAAC) depends on this. Do not subnet a LAN smaller than /64.
• Point-to-point exceptions. /127 is permitted on router-to-router links (RFC 6164), and /128 is used for loopbacks.
• No masks. IPv6 uses prefix length only, never dotted masks.

IPv6 Prefix Reference

Splitting a larger prefix into smaller ones follows $2^{,p_{\text{new}} - p_{\text{old}}}$. A /48 contains $2^{64-48} = 2^{16} = 65{,}536$ subnets of size /64.

IPv6 Text Representation (RFC 5952)

The calculator normalizes every IPv6 input to the canonical form from RFC 5952:

• hexadecimal digits are lowercase;
• leading zeros in each hextet are removed;
• the longest run of consecutive all-zero hextets (length two or more) is replaced with ::, and if there is a tie the first run is compressed;
• a single zero hextet is not compressed.

So 2001:0DB8:ABCD:0012:0000:0000:0000:0001 becomes 2001:db8:abcd:12::1.

Subnet Bit Calculator: Borrowing Bits and Splitting Networks

Subnetting a block means borrowing host bits to make network bits. If you borrow $b$ bits from a block, you create:

each with $h - b$ host bits. Moving from /24 to /27 borrows $b = 3$ bits, producing $2^{3} = 8$ subnets of $2^{5} = 32$ addresses each (30 usable).

Choosing a Prefix for a Host Count

To fit $n$ hosts into a conventional IPv4 subnet, find the smallest $h$ satisfying:

Then the prefix is $p = 32 - h$. For 100 hosts: $\log_{2}(102) \approx 6.67$, so $h = 7$ and $p = 25$ (126 usable). For 500 hosts: $\log_{2}(502) \approx 8.97$, so $h = 9$ and $p = 23$ (510 usable).

VLSM Example

Variable Length Subnet Masking allocates different prefix sizes from one parent block. Allocate largest first to keep blocks aligned. Plan for 192.168.10.0/24 with these requirements:

The remaining space (192.168.10.228 – 192.168.10.255) is free for growth. Paste any network address from the table into the calculator to verify its boundaries.

Special-Purpose and Private IPv4 Ranges

For IPv6, 2001:db8::/32 is reserved for documentation (RFC 3849), fc00::/7 for unique local addresses (RFC 4193), and fe80::/10 for link-local.

Common Subnetting Mistakes

• Applying "minus 2" to /31 and /32. Both are valid with RFC 3021 and host-route semantics, and the calculator reports them accordingly.
• Misaligned subnets. A subnet must start on a multiple of its block size. 10.0.0.100/26 is not a network address; the network is 10.0.0.64/26.
• Confusing wildcard and subnet masks. 0.0.0.255 is a wildcard (matches any last octet); 255.255.255.0 is the netmask for the same /24.
• Entering CIDR notation in the address field. Put 192.168.1.10 in the address field and 24 in the prefix field. Include no slash, port, or brackets in the address.
• Subnetting IPv6 LANs smaller than /64. This breaks SLAAC and some neighbor discovery behaviour.
• Leading zeros. 192.168.001.010 is rejected: some systems read leading-zero octets as octal, so the calculator requires unambiguous decimal.

Limitations and Standards

This is a pure mathematical subnet calculator. It follows:

• RFC 4632 – CIDR addressing and prefix notation
• RFC 3021 – 31-bit prefixes on IPv4 point-to-point links
• RFC 4291 – IPv6 addressing architecture (no broadcast; multicast and anycast)
• RFC 5952 – Recommended IPv6 text representation

It does not scan networks, perform DNS lookups, check public reachability, or call any external API. It also does not subtract cloud-provider reservations. AWS and Azure reserve five addresses in every subnet and Google Cloud reserves four in each primary range, so subtract those from the conventional host count when sizing cloud subnets and confirm current figures in your provider's documentation.

Frequently Asked Questions