Free IPv6 Calculator

IPv6 Subnet Calculator

Enter any IPv6 address and prefix length to instantly calculate the network prefix, network address, first and last address, total address count, number of /64 subnets, address type, and a hex nibble visualizer showing which bits are the network portion. Includes a complete IPv6 subnetting guide, prefix reference table, and worked examples. Runs entirely in your browser — no data sent anywhere.

Auto-calculates as you type Hex nibble visualizer Subnet splitter Complete prefix guide Browser-only Always free
Network tool
Enable JavaScript to run lookups and interactive features on this page.

Hero, guides, and sidebar links below work without JavaScript. The interactive checker needs JavaScript enabled in your browser.

What Is IPv6 Subnetting?

IPv6 subnetting divides an IPv6 address space into smaller network segments using a prefix length — a number from /0 to /128 that specifies how many of the 128 address bits are the network portion. IPv6 subnetting follows the same fundamental principles as IPv4 subnetting, but operates at a vastly different scale: where IPv4 has 4.3 billion total addresses, IPv6 has 340 undecillion (3.4 × 10³⁸) — a number so large that every atom on Earth could have its own IPv6 address.

IPv6 subnet calculator — calculate network prefix, address range and subnets for any IPv6 address

IPv6's 128-bit address space is hierarchically structured — from the global routing prefix assigned by an ISP down to individual /64 subnets and interface IDs

The key practical difference from IPv4: in IPv6, the standard subnet size for end-user networks is /64 — each /64 contains 2⁶⁴ = 18.4 quintillion host addresses. Instead of carefully rationing addresses like IPv4, IPv6 design intentionally allocates large blocks — organisations typically receive a /48 (65,536 /64 subnets) from their ISP, and home users receive a /56 (256 /64 subnets). The abundance is by design.

Why /64 is always the interface subnet: IPv6 uses SLAAC (Stateless Address Autoconfiguration) for automatic address assignment. SLAAC constructs a device's IP address by combining the /64 network prefix with the device's 64-bit interface identifier (derived from the MAC address or a random value). For SLAAC to work, the subnet must always be exactly /64 — subnets smaller than /64 break SLAAC and are therefore never used for interface subnets in practice.

IPv6 Address Structure — How 128 Bits Are Organised

An IPv6 address is 128 bits, written as eight groups of four hexadecimal digits separated by colons. Each group (called a "hextet" or "group") represents 16 bits. Understanding the structure of these 128 bits is the foundation of IPv6 subnetting:

Full format: 2001:0db8:85a3:0000:0000:8a2e:0370:7334
Compressed: 2001:db8:85a3::8a2e:370:7334
(:: replaces one or more consecutive groups of all-zeros)

For 2001:db8::/32 allocation:

|<─── 32 bits ──────>|<── 16 ──>|<── 16 ──>|<─────── 64 bits (interface ID) ──────────>|
2001:0db8 : xxxx : xxxx : xxxx:xxxx:xxxx:xxxx
Global routing Site Subnet Interface identifier
prefix prefix (admin) (device address)

A typical /48 site allocation broken down:
2001:db8:1234: [16-bit subnet field] : [64-bit interface ID]
|<───────── 48 bits (ISP assigns) ────>|<─16─>|<─── 64 bits ───>|
subnet interface addr
→ 2^16 = 65,536 possible /64 subnets in one /48

The Three Parts of an IPv6 Unicast Address

Most global IPv6 unicast addresses (the type used for internet-routable hosts) have three functional sections:

PartBitsDescriptionExample
Global Routing PrefixTypically 48 bitsAssigned by APNIC/RIPE to ISPs, and by ISPs to customers. Identifies the organisation globally. All traffic to this prefix routes to the same organisation.2001:db8:1234::/48
Subnet IDTypically 16 bitsChosen by the organisation's network administrator to identify subnets within their allocation. 16 bits = 65,536 possible subnets from a /48. Jio, Airtel, and other ISPs allocate subnet IDs to customers within their blocks.:0001: through :ffff:
Interface ID64 bitsIdentifies the specific device interface within the subnet. Generated automatically via SLAAC (from MAC address via EUI-64 or random privacy extensions) or assigned manually. Always 64 bits.::1 (loopback), random, or EUI-64

IPv6 Prefix Length Reference — /0 to /128

Every IPv6 prefix length serves a specific purpose in the allocation hierarchy. Here is the complete reference for the prefix lengths you will encounter in real-world IPv6 deployments, RIR allocations, and network design:

PrefixTotal AddressesTypical AllocationWho Receives ItNotes
/0All 2¹²⁸Default routeInternet routing::/0 is the default route — matches all IPv6 traffic. Used in routing tables.
/122¹¹⁶RIR allocationAPNIC, RIPE, ARINRegional Internet Registries receive /12 blocks from IANA to allocate to ISPs.
/232¹⁰⁵ISP allocationLarge ISPs (Jio, Airtel)ISPs receive /23 or larger blocks from APNIC to allocate to their customers.
/322⁹⁶ ≈ 7.9×10²⁸ISP allocation unitISPs and large organisationsStandard allocation from APNIC to ISPs. Jio, Airtel, BSNL have /32 or larger allocations. Contains 65,536 /48 customer sites.
/402⁸⁸ ≈ 3.1×10²⁶ISP customer blockHosting providersHosting providers often assign /40 blocks to data centre customers. Contains 256 /48 site allocations.
/482⁸⁰ ≈ 1.2×10²⁴Site allocationOrganisations, enterprisesStandard allocation for a single organisation site. Contains 65,536 /64 subnets — sufficient for any enterprise network. IETF recommends /48 per site.
/522⁷⁶ ≈ 7.5×10²²Customer siteSMB organisationsSome ISPs allocate /52 to smaller business customers. Contains 4,096 /64 subnets.
/562⁷² ≈ 4.7×10²¹Home/residentialHome users, small businessesStandard home broadband IPv6 allocation in many countries. Contains 256 /64 subnets — 256 home network segments for one household.
/602⁶⁸ ≈ 2.9×10²⁰Small home allocationSome residential ISPsSome ISPs (including some Jio deployments) allocate /60 to home users. Contains 16 /64 subnets.
/642⁶⁴ ≈ 1.8×10¹⁹Single subnetIndividual LAN segmentsThe ALWAYS-required subnet size for SLAAC autoconfiguration. Every LAN, VLAN, and server network is a /64. Never subnet smaller than /64 for user-facing networks.
/802⁴⁸ ≈ 2.8×10¹⁴Rarely usedSpecial applicationsTechnically valid but breaks SLAAC. Only used in very specific non-standard configurations.
/1122¹⁶ = 65,536Point-to-multipointSpecial useProvides IPv4-equivalent /16 address space. Rarely needed due to IPv6 abundance.
/120256IPv4-equivalent /24Migration scenariosComparable in size to an IPv4 /24. Used in IPv4-to-IPv6 migration planning for familiarity.
/1264Point-to-point linksRouter interconnectsIPv6 equivalent of IPv4 /30 — used for router-to-router point-to-point links. 2 usable addresses for routers.
/1272Point-to-point (RFC 6164)Router interconnectsRecommended by RFC 6164 for point-to-point router links. No reserved network/broadcast addresses — both usable.
/1281Single host routeLoopback, anycast, specific hosts::1/128 is the loopback address. Used in routing tables for specific host routes, anycast addresses, and firewall rules matching one IP.

IPv6 Address Types — Global, ULA, Link-Local, Multicast

IPv6 eliminates the concept of private vs public IP addresses from IPv4 and replaces it with distinct address types that serve different communication scopes. Understanding these types is essential for IPv6 network design:

Global Unicast (GUA) 2000::/3
Internet-routable addresses. Start with binary 001 (hex 2 or 3). These are the IPv6 equivalent of public IPv4 addresses. Assigned by RIRs (APNIC for Asia-Pacific) through ISPs to end users. Example: 2001:db8::/32 is reserved for documentation.
Unique Local Address (ULA) fc00::/7
IPv6's equivalent of RFC 1918 private addresses. Start with fc00:: or fd00::. Not routable on the public internet. The fd00::/8 range uses a pseudo-random 40-bit global ID to minimise conflicts when merging networks. Used for internal-only addressing in corporate networks.
Link-Local fe80::/10
Automatically configured on every IPv6-enabled interface. Scope is the local network segment only — never forwarded by routers. Used for neighbor discovery, router solicitation, and DHCPv6. Every IPv6-capable device always has a link-local address, even if no other IPv6 is configured.
Multicast ff00::/8
IPv6 has no broadcast — multicast replaces it entirely. ff02::1 is "all nodes on link," ff02::2 is "all routers on link." Multicast scope is encoded in the address: ff02::/16 = link-local, ff05::/16 = site-local, ff0e::/16 = global. NDP (Neighbor Discovery Protocol) uses multicast exclusively.
Loopback ::1/128
::1 is the IPv6 loopback address (equivalent to 127.0.0.1). Traffic sent to ::1 stays on the local device. Only one loopback address exists in IPv6 (compared to an entire /8 in IPv4). Always written as ::1 in compressed notation.
Documentation 2001:db8::/32
2001:db8::/32 is reserved by RFC 3849 for use in documentation, examples, and technical writing. Never assigned to live networks. All examples in RFCs, textbooks, and this page use 2001:db8:: addresses. Equivalent to 192.0.2.0/24 in IPv4.

No More Broadcast in IPv6

IPv6 completely eliminates broadcast — there is no IPv6 broadcast address. All functions that used broadcast in IPv4 (ARP, DHCP discovery, routing protocol hellos) are replaced by multicast in IPv6. This eliminates broadcast storms and significantly reduces the noise on large layer-2 segments. The Neighbor Discovery Protocol (NDP) uses multicast instead of ARP for address resolution.

IPv6 Subnetting Worked Examples

Three complete examples covering ISP allocation breakdown, enterprise site design, and the /127 point-to-point link — the most important IPv6 subnetting scenarios in real networks:

Example 1: Break a /48 into /64 subnets

Given: 2001:db8:1234::/48 — divide into /64 subnets
1
/48 to /64 = 16 additional bits borrowed. Number of subnets = 2¹⁶ = 65,536 /64 subnets
2
Each /64 subnet has: 2⁶⁴ = 18,446,744,073,709,551,616 host addresses — more than enough for any LAN
3
First subnet: 2001:db8:1234:0000::/64 (hosts: 2001:db8:1234:0000::1 to 2001:db8:1234:0000:ffff:ffff:ffff:ffff)
4
Second subnet: 2001:db8:1234:0001::/64
5
Last subnet: 2001:db8:1234:ffff::/64
6
Pattern: Increment the 4th hextet (subnet field) from 0000 to ffff — 65,536 subnets total

Example 2: Enterprise Network Planning from a /48

Design: Assign subnets from 2001:db8:acme::/48
1
Floor 1 servers: 2001:db8:acme:0001::/64 (subnet ID = 0x0001)
2
Floor 1 workstations: 2001:db8:acme:0002::/64 (subnet ID = 0x0002)
3
Floor 2 servers: 2001:db8:acme:0011::/64 (use hex prefix to organise by floor)
4
DMZ: 2001:db8:acme:0f00::/64 (use 0xF prefix for DMZ subnets)
5
VoIP: 2001:db8:acme:a000::/64 (use 0xA prefix for voice)
6
Result: 65,535 subnets still available for expansion — plan the 16-bit subnet field hierarchically using the high nibbles for building/function and low nibbles for segment number

Example 3: IPv6 Point-to-Point Link (/127)

WAN link between Router A and Router B
1
Assign a /127 from your infrastructure range: 2001:db8:acme:ffff::/127
2
Router A interface: 2001:db8:acme:ffff::0/127 (the ::0 address)
3
Router B interface: 2001:db8:acme:ffff::1/127 (the ::1 address)
4
Why /127 not /126? RFC 6164 recommends /127 for point-to-point links — both addresses usable (no network/broadcast reserved). /126 wastes 2 addresses unnecessarily.
5
Security note: /127 also prevents "subnet-router anycast address" attacks that are possible with /126 and larger subnets.

IPv6 vs IPv4 Subnetting — Key Differences

IPv6 subnetting uses the same fundamental prefix-length mechanism as IPv4 CIDR but differs significantly in scale, conventions, and features:

AspectIPv4IPv6
Address length32 bits128 bits
Total addresses~4.3 billion~340 undecillion (3.4×10³⁸)
Standard host subnet/24 (254 hosts)/64 (18.4 quintillion hosts)
NotationDecimal dotted (192.168.1.0/24)Hexadecimal colon-separated (2001:db8::/32)
Subnet mask255.255.255.0 or /24Prefix length only — no subnet mask notation
Wildcard mask0.0.0.255 (for ACLs)Not used — prefix length is universal
BroadcastYes (subnet broadcast address)No — replaced entirely by multicast
Private addressesRFC 1918 (10.x, 172.16.x, 192.168.x)ULA (fc00::/7, typically fd00::)
Loopback127.0.0.1 (entire /8)::1 (single address /128)
Auto-configurationDHCP (stateful only)SLAAC (stateless, /64 required) + DHCPv6
P2P link size/30 (4 addresses, 2 usable)/127 (2 addresses, both usable per RFC 6164)
Reserved/network addressFirst address in each subnetNo reserved network address — first address usable

Who Uses an IPv6 Subnet Calculator — 8 Real-World Scenarios

Cloud VPC Design
AWS, Azure, and GCP all support IPv6 VPCs. Each cloud provider assigns a /56 or /48 prefix to the VPC. The calculator helps plan how to divide this into /64 subnets for each availability zone and service tier.
Enterprise Network Planning
Designing IPv6 for a multi-site organisation requires systematically allocating /48 blocks per site and /64 subnets per VLAN. The 16-bit subnet field in a /48 gives 65,536 subnets — enough for any enterprise hierarchy.
Home ISP Allocation
Jio FTTH and Airtel Xstream Fiber are beginning to assign IPv6 prefixes (/56 or /60) to home routers. Understanding the allocation lets you plan internal VLANs for IoT, work, and media devices on separate /64 subnets.
CCNA / CCNP IPv6 Study
Cisco certifications heavily test IPv6 subnetting. The calculator verifies practice question answers and helps build intuition for /48, /56, /64, /126, /127, and /128 calculations — the prefixes most commonly tested.
Router Interface Configuration
Assigning IPv6 addresses to router interfaces requires knowing the correct prefix. The calculator confirms that a /127 is correctly sized for point-to-point links, and verifies that /64 is being used for LAN interfaces where SLAAC is required.
Application and API Development
Developers implementing IPv6 address validation, network range checks, or geolocation systems need to verify IPv6 subnet arithmetic. The calculator provides reference outputs for testing code that handles IPv6 prefixes.
Firewall Rule Writing
Writing IPv6 ACLs and security group rules requires knowing the exact network prefix. AWS Security Groups, Linux ip6tables, and Cisco IOS all use IPv6 CIDR notation. The calculator confirms prefixes before committing rules to production.
IPv4 to IPv6 Migration
Planning a dual-stack migration requires mapping existing IPv4 subnets to equivalent IPv6 /64 subnets. The calculator helps understand scale differences and plan the IPv6 addressing scheme that mirrors your existing IPv4 topology.

For more tools see subnet calculator (IPv4), CIDR calculator, and IP range calculator.

Frequently Asked Questions — IPv6 Subnet Calculator

Why is /64 the standard IPv6 subnet size?

The /64 subnet size is required for SLAAC (Stateless Address Autoconfiguration) — the mechanism that allows IPv6 devices to automatically configure their own IP addresses without a DHCP server. SLAAC works by combining the /64 network prefix with a 64-bit interface identifier derived from the device's MAC address (via EUI-64) or generated randomly (RFC 7217). If the subnet is smaller than /64, there aren't 64 bits for the interface ID and SLAAC cannot function. Since SLAAC is fundamental to IPv6's plug-and-play design, /64 became the mandatory size for all user-facing subnets. Every LAN segment, VLAN, and server network uses exactly /64.

What is the difference between /48, /56, and /64 in IPv6?

/48, /56, and /64 represent different levels in the IPv6 allocation hierarchy: /48 is a "site" allocation — typically assigned by an ISP to an organisation, providing 65,536 /64 subnets. Large enterprises and data centres receive a /48 (or larger) to address their entire infrastructure. /56 is a "home/residential" allocation — many ISPs assign a /56 to home broadband subscribers, providing 256 /64 subnets — more than enough for home network segmentation (IoT, guest, work, etc.). /64 is a single subnet — every individual network segment (LAN, VLAN, Wi-Fi network) is exactly /64. You never manually divide a /64 into smaller pieces for host addressing.

How do I read and expand IPv6 addresses?

IPv6 addresses have two simplification rules: (1) Leading zeros in each group can be omitted — 0db8 can be written as db8. (2) One or more consecutive groups of all-zeros can be replaced with :: — but only once per address. To expand: replace :: with the number of 0000 groups needed to make 8 groups total. For 2001:db8::1: count existing groups (2001, db8, 1 = 3 groups), need 8 total, so :: = 5 groups of zeros: 2001:0db8:0000:0000:0000:0000:0000:0001. The /prefix tells you how many of the leftmost bits are the network portion.

Does IPv6 have broadcast addresses?

No. IPv6 completely eliminates broadcast. Every function that used broadcast in IPv4 is replaced by multicast in IPv6: ARP is replaced by NDP (Neighbor Discovery Protocol) using solicited-node multicast addresses (ff02::1:ff00:0/104), DHCP discovery uses ff02::1:2 (all-DHCP-relay-agents-and-servers), routing protocol hellos use ff02::5 (OSPF all-routers) and ff02::6 (OSPF designated routers), and general "all hosts" is ff02::1. This eliminates broadcast storms and significantly improves performance on large layer-2 segments.

What is a ULA (Unique Local Address) and when should I use it?

ULA (Unique Local Address, RFC 4193) is the IPv6 equivalent of RFC 1918 private addresses (10.x, 172.16.x, 192.168.x). ULAs start with fc00::/7, with fd00::/8 being the commonly used range. They are not routable on the public internet. Use ULA when: you want private IPv6 addressing for internal services that should never be internet-accessible, you're on a network that doesn't have a globally assigned IPv6 prefix from an ISP, or you're building lab/test infrastructure. For production internet-accessible services, use globally assigned unicast addresses (2000::/3) from your ISP. ULA fd00:: addresses include a pseudo-random 40-bit global ID to minimise conflicts when merging two ULA networks.

How do Jio and Airtel deploy IPv6 in India?

Jio (Reliance Jio, ASN 55836) and Airtel (Bharti Airtel, ASN 9498) both have IPv6 allocations from APNIC. Jio's FTTH (JioFiber) and Jio 5G connections increasingly provide IPv6 by default, typically assigning /56 or /60 prefixes to customer routers via DHCPv6-PD (Prefix Delegation). Jio's mobile 4G/5G network uses IPv6 natively with dual-stack. Airtel Xstream Fiber similarly provides IPv6, often assigning a /56 or /48 via DHCPv6-PD. If your Jio or Airtel connection provides IPv6, your router receives a prefix delegation and automatically assigns /64 subnets to each of your home network interfaces. Check your router's IPv6 settings to see your assigned prefix.

What is EUI-64 and how does it relate to IPv6 subnetting?

EUI-64 is the method SLAAC uses to generate the 64-bit interface identifier portion of an IPv6 address from a device's 48-bit MAC address. The process: (1) Split the MAC address in the middle and insert ff:fe (e.g., MAC 00:1A:2B:3C:4D:5E becomes 00:1A:2B:ff:fe:3C:4D:5E). (2) Flip the 7th bit (Universal/Local bit) of the first byte. (3) The resulting 64 bits become the interface ID appended to the /64 prefix. For privacy, RFC 7217 and RFC 4941 define "privacy extensions" that use a random 64-bit interface identifier instead of the MAC-derived EUI-64, preventing device tracking. Modern operating systems (Windows, Linux, macOS, Android, iOS) use privacy extensions by default.

What is DHCPv6-PD (Prefix Delegation) and how does it work?

DHCPv6-PD (Prefix Delegation, RFC 3633) is the mechanism by which your ISP (Jio, Airtel, etc.) automatically assigns a block of IPv6 addresses to your home router. Instead of just giving your router a single IP address, DHCPv6-PD gives it an entire prefix (e.g., a /56) which the router then subdivides into /64 subnets and assigns to each of its internal interfaces (LAN, Wi-Fi 2.4GHz, Wi-Fi 5GHz, guest network, etc.). The process: (1) Router sends DHCPv6 Solicit with IA_PD option requesting a prefix. (2) ISP DHCPv6 server responds with a /56 (or /60) delegation. (3) Router allocates /64 subnets from this delegation to each interface. (4) Devices on each /64 use SLAAC to self-configure their addresses from the /64 prefix.

Is this IPv6 subnet calculator free?

Yes — completely free, no signup, no account required. All calculations run entirely in your browser using JavaScript BigInt operations for 128-bit precision — no IPv6 address data is transmitted to any server. The calculator supports all IPv6 address formats (full, compressed with ::, mixed IPv4-mapped), all prefix lengths from /0 to /128, and includes a hex nibble visualizer, interactive subnet splitter, preset buttons for common use cases, and copy-per-field functionality.

Related IP Calculators & Network Tools

Advertisement