IPv4 has been the internet's primary addressing scheme since the 1980s, and its dotted-decimal format — four numbers separated by dots, like 203.0.113.42 — is probably the most recognizable piece of networking notation there is.
Reading the format
Each of the four numbers (called an octet) represents 8 bits and ranges from 0 to 255, for a total 32-bit address. That gives IPv4 about 4.3 billion possible unique addresses — an enormous number in 1981 when the standard was finalized, and nowhere near enough for a world with billions of internet-connected phones, laptops, and IoT devices today.
Why scarcity shaped everything else
IPv4 address exhaustion — the regional internet registries running out of new blocks to allocate — is the direct reason two other things exist: NAT and private addressing, which let millions of devices share a small number of public addresses, and IPv6, which was designed specifically to make scarcity a non-issue with a vastly larger address space. See IPv4 vs IPv6 for how the two now coexist.
Why IPv4 is still everywhere
Despite IPv6 existing for decades, IPv4 remains dominant in day-to-day browsing because so much existing infrastructure — routers, ISPs, corporate networks — was built around it, and the transition only makes sense gradually. Most connections today are dual-stack, running both protocols side by side, with IPv4 often still the default for many applications.
FAQ
Why can't we just make more IPv4 addresses?
The 32-bit address size is baked into the protocol itself — every piece of internet infrastructure that handles IPv4 traffic expects exactly four 8-bit numbers. Expanding it would mean a fundamentally different protocol, which is exactly what IPv6 is.
What's the difference between an IPv4 address and a subnet?
A single IPv4 address identifies one device (or, behind NAT, one shared exit point). A subnet is a range of addresses grouped together, usually written with a prefix like /24, that a network administrator manages as a block.