Comparison Guide

IPv4 vs IPv6

The internet ran out of IPv4 addresses. Here's how the world is moving to IPv6 — and what it means for you.

12 min readTechnical guideUpdated 2026

Why there are two versions of the Internet Protocol

The internet was designed in an era when billions of connected devices seemed unimaginable. IPv4, the original addressing system, was engineered to support approximately 4.3 billion unique addresses. That number felt enormous in 1981 — it doesn't today.

By the late 1990s, engineers recognized that the IPv4 address pool would eventually be exhausted. Rather than patch the existing system, they designed a new protocol from scratch: IPv6. The goal was not just to create more addresses, but to build a better protocol for the modern internet — with improved security, simpler routing, and automatic configuration built in.

Today, the internet runs on both protocols simultaneously. Most modern devices support both, and many networks operate in "dual-stack" mode — preferring IPv6 when available and falling back to IPv4 when not. The transition is ongoing and has taken far longer than anticipated.

The history of IPv4

IPv4 was defined in September 1981 in RFC 791, authored by Vint Cerf and Bob Kahn — often called "the fathers of the internet." It replaced a patchwork of earlier addressing schemes used on ARPANET and introduced the 32-bit address format that remains in wide use today.

The 32-bit address space theoretically provides 4,294,967,296 unique addresses. In practice, the usable number is smaller because large blocks were reserved for special purposes, and early allocation policies were often wasteful. For example, MIT was given the entire 18.0.0.0/8 block — 16.7 million addresses — for a network that never needed anywhere near that many.

The internet grew explosively through the 1990s as it opened to commercial and consumer use. Workarounds like NAT (Network Address Translation) extended IPv4's lifespan by allowing millions of devices to share a single public address, but this came at the cost of added complexity and limitations for peer-to-peer applications.

On February 3, 2011, IANA (the Internet Assigned Numbers Authority) allocated its last five IPv4 /8 blocks to regional registries. By 2015, most regional registries had exhausted their free pools. IPv4 addresses are now scarce, expensive, and actively traded on secondary markets — some /24 blocks (256 addresses) sell for thousands of dollars.

Understanding the IPv4 address format

An IPv4 address consists of 32 binary bits, typically written as four decimal numbers (called octets) separated by dots. Each octet represents 8 bits, so each can range from 0 to 255.

192.168.1.100
11000000 . 10101000 . 00000001 . 01100100
↑ 192       ↑ 168       ↑ 1         ↑ 100

An IPv4 address is split into two parts: the network portion (which identifies the network) and the host portion (which identifies the specific device on that network). The split is defined by a subnet mask or CIDR notation.

CIDR notation

192.168.1.0/24 → 254 hosts
10.0.0.0/8 → 16.7M hosts
203.0.113.0/30 → 2 hosts

The /24 means the first 24 bits are the network address, leaving 8 bits for hosts.

Address classes (historical)

Class A: 1.0.0.0 – 126.255.255.255
Class B: 128.0.0.0 – 191.255.255.255
Class C: 192.0.0.0 – 223.255.255.255

Classful addressing was replaced by CIDR in 1993 for more efficient allocation.

Understanding the IPv6 address format

IPv6 uses 128-bit addresses, written as eight groups of four hexadecimal digits, separated by colons. This produces addresses that look intimidating at first, but there are rules for compressing them.

Full form:
2001:0db8:85a3:0000:0000:8a2e:0370:7334
Compressed form (leading zeros dropped, consecutive zero groups collapsed):
2001:db8:85a3::8a2e:370:7334

Compression rules

  • Leading zeros in each group can be omitted (0042 → 42)
  • One or more consecutive all-zero groups can be replaced with ::
  • The :: shorthand can only appear once in an address

Notable IPv6 addresses

::1 → Loopback (like 127.0.0.1)
:: → Unspecified address
fe80::/10 → Link-local
fc00::/7 → Unique local (private)
2001:db8::/32 → Documentation

Side-by-side comparison

FeatureIPv4IPv6
Address length32 bits128 bits
Total addresses~4.3 billion340 undecillion
Address formatDotted decimal (192.168.1.1)Colon hex (2001:db8::1)
SecurityIPsec optionalIPsec designed-in
Auto-configurationRequires DHCP or manual setupSLAAC — no server needed
Header sizeVariable (20–60 bytes)Fixed 40 bytes
NAT required?Yes, at scaleNo — each device gets unique global IP
FragmentationBy routers and hostsBy hosts only (faster routing)
QoS supportToS field (limited)Traffic class + flow label
MulticastOptionalMandatory — replaces broadcast
ChecksumHeader checksumNo header checksum (handled by transport layer)
Mobile supportLimitedNative Mobile IPv6 (MIPv6)

Why IPv6 is technically superior for modern routing

IPv6 was not just designed to add more addresses — it was engineered to fix fundamental inefficiencies in IPv4 that had accumulated over decades of workarounds.

Simplified routing tables

IPv6's hierarchical address structure allows ISPs to aggregate routes more efficiently. Instead of advertising thousands of small prefixes, providers can advertise one large block. This reduces the size of global routing tables and improves router performance.

No more NAT complexity

NAT breaks the end-to-end connectivity principle of the internet — every device behind NAT must work around it to host services or receive inbound connections. IPv6's enormous address space eliminates the need for NAT, restoring direct device-to-device communication.

Stateless address autoconfiguration

IPv6 devices can configure their own global addresses without a DHCP server using SLAAC (Stateless Address Autoconfiguration). A device uses its network prefix (from the router) and its own MAC address to generate a unique global address automatically.

Fixed header size

IPv4 headers can vary in size because of optional fields, requiring routers to parse them carefully. IPv6 uses a fixed 40-byte header, making router processing faster and more predictable. Optional data is handled through extension headers that only appear when needed.

How the transition from IPv4 to IPv6 works

The internet cannot switch from IPv4 to IPv6 overnight — billions of devices, applications, and services need to be updated over time. Engineers developed several transition mechanisms to allow IPv4 and IPv6 to coexist during this period.

Dual-stack

Devices and networks run both IPv4 and IPv6 simultaneously. When both are available, modern devices prefer IPv6 using a process called "Happy Eyeballs" (RFC 8305), which races both connection types and uses whichever responds first. This is the most common approach today.

Tunneling

IPv6 packets are encapsulated inside IPv4 packets to cross IPv4-only network segments. Techniques include 6to4, 6in4, and Teredo. These add overhead and complexity and are generally considered stopgap measures rather than long-term solutions.

NAT64 and DNS64

Allows IPv6-only clients to reach IPv4 servers. NAT64 translates between IPv6 and IPv4 packets at the network boundary. DNS64 synthesizes AAAA records for IPv4-only servers, making them reachable from IPv6-only networks. Common in modern mobile networks.

464XLAT

A modern technique used by mobile carriers to provide IPv4 connectivity over IPv6-only networks. The mobile device runs a small CLAT (client-side translator) that converts local IPv4 traffic to IPv6 for transit, then NAT64 converts back to IPv4 at the carrier's edge.

IPv6 adoption around the world

IPv6 adoption has been growing steadily but unevenly across regions and sectors. Mobile networks have been faster adopters than fixed broadband, driven by NAT64/464XLAT making IPv6-only deployments feasible.

45%+
Google users reaching Google over IPv6 (2025)
50%+
US mobile traffic on IPv6 (Verizon, T-Mobile lead)
95%+
IPv6 adoption in India (driven by Reliance Jio)

Adoption is highest in countries where ISPs built new infrastructure from the ground up (India, some African markets) rather than upgrading legacy systems. Enterprise adoption lags behind consumer ISPs due to the complexity of migrating existing infrastructure, security policies, and applications.

What IPv6 means for you today

For most internet users, the IPv4 to IPv6 transition is invisible. Your device handles it automatically, and the websites you visit will work whether they use IPv4, IPv6, or both. But there are a few practical implications worth understanding.

Checking if you have IPv6

Your current IP lookup will show whether your connection has an IPv6 address. If you see both IPv4 and IPv6 addresses, your network is dual-stack. If you only see IPv4, your ISP hasn't yet deployed IPv6 to your area.

Check your IP now →

Privacy considerations

IPv6 addresses can expose your device's MAC address if using EUI-64 format. Modern systems use privacy extensions (RFC 8981) to generate temporary, randomly changing IPv6 addresses for outgoing connections — check that your OS has this enabled.

For developers and sysadmins

Applications must be tested with both protocols. Network rules, firewall configurations, and logging systems all need to handle IPv6. Assuming IPv4 addresses fit in specific field sizes or formats is a common source of bugs.

VPNs and IPv6 leaks

Some VPNs only tunnel IPv4 traffic. If your connection has IPv6 and the VPN doesn't handle it, your real IPv6 address can leak to websites you visit — revealing your identity even through a VPN. Always check your VPN's IPv6 leak protection.

Common questions

Should I be worried if my connection doesn't have IPv6?

Not immediately. IPv4 still works for virtually everything, and NAT has extended its effective lifespan significantly. However, IPv6-only services are still rare, so lack of IPv6 support won't cause problems for most users today. Your ISP will eventually deploy IPv6 when they upgrade their infrastructure.

Is IPv6 faster than IPv4?

In theory, yes — IPv6 has a simpler header structure that requires less processing by routers, and it eliminates NAT latency. In practice, the difference is typically small and often unnoticeable. What matters more is your ISP's network quality and the server's location. Some users report slightly faster page loads with IPv6 due to preferred routing paths.

Is there an IPv5?

IPv5 existed as an experimental real-time streaming protocol (ST2, RFC 1190) that was never widely deployed. When the next version of IP was standardized, it jumped to IPv6 to avoid confusion with the experimental IPv5.

Will we ever run out of IPv6 addresses?

340 undecillion (3.4 × 10³⁸) addresses is a practically inexhaustible supply at any foreseeable scale. Even if every atom on the surface of the Earth were a networked device, IPv6 would still have addresses to spare. The concern about running out of IPv6 addresses is not considered realistic.

Why hasn't IPv6 fully replaced IPv4 yet?

The transition is slower than anticipated because IPv4 still works — thanks to NAT — and upgrading infrastructure is expensive. Enterprises must update firewalls, load balancers, monitoring tools, and applications. Many ISPs have delayed because customers won't notice the difference, reducing the business incentive to invest. The transition is happening, just gradually.

Key takeaways

  • IPv4 uses 32-bit addresses; IPv6 uses 128-bit addresses providing 340 undecillion unique IPs.
  • IANA exhausted the IPv4 address pool in 2011; most regional registries followed shortly after.
  • IPv6 eliminates the need for NAT, simplifies routing, and includes built-in security features.
  • Most modern devices and networks support both protocols via dual-stack operation.
  • Transition mechanisms like NAT64 and tunneling bridge the gap during the migration period.
  • Check that your VPN handles IPv6 to avoid privacy-exposing leaks.