What Is an IP Address?

An Internet Protocol address is a routing label. When one device sends an IP packet, the source and destination addresses give networks enough information to move that packet toward the right connection. That job is fundamental, but narrower than the word “address” may suggest: an IP address is not a permanent name for a person, a precise map coordinate, or proof of who used a device.

Why IP addresses exist

The Internet joins many independently operated networks. Routers examine the destination prefix of a packet and choose a next hop; they do not need a complete map of every individual device. Replies use the source information, sometimes after an intermediary has rewritten it. Applications such as browsers usually work with domain names, but DNS ultimately supplies addresses that the networking stack can use.

An address belongs to an interface in a particular network context. A laptop can have a private Wi-Fi address, a temporary IPv6 address, and traffic that appears externally under a router’s public IPv4 address. Disconnecting, changing networks, restarting a gateway, or a provider’s lease policy can change what is assigned.

IPv4 and IPv6

IPv4 writes four eight-bit numbers in decimal. Its limited address space helped make address sharing through NAT widespread. IPv6 writes eight groups of hexadecimal values, allows leading zeros to be omitted, and permits one run of zero groups to be compressed as ::. Compression changes the notation, not the numeric address. You can use the IP Address Checker to normalize either version and inspect special classifications.

Public, private, and special-purpose ranges

A globally routable address can be announced across the public Internet. Private IPv4 ranges—10.0.0.0/8, 172.16.0.0/12, and 192.168.0.0/16—are intended for local networks and can be reused by unrelated organizations. IPv6 has unique-local space for a broadly similar private use. Neither kind should be treated as a public destination.

“Not private” still does not automatically mean globally routable. Loopback addresses refer to the local host; link-local addresses work only on a local link; multicast addresses identify groups; unspecified addresses mean no particular interface; and documentation ranges are reserved for examples. Context matters when interpreting a library’s flags.

Dynamic and static assignments

A static address is expected to remain assigned under an administrative arrangement. A dynamic address comes from a pool or lease and may change. These labels describe assignment practice, not whether an address is public. A home router can retain the same dynamic public address for months, while a hosted server can be renumbered despite a stable service name. DNS gives services an abstraction that can continue while addresses change.

What NAT changes

Network Address Translation commonly lets many private IPv4 devices share a smaller number of public addresses. A gateway rewrites source addresses and ports on outbound flows, remembers the mapping, and reverses it for replies. Carrier-grade NAT can add another sharing layer inside an ISP. Consequently, the public source observed by a website may represent a household, office, mobile gateway, VPN exit, or large provider pool.

NAT is not the same as a firewall, although consumer routers often perform both functions. It also does not make activity anonymous: websites can use accounts, cookies and other signals, while providers may retain allocation records. IPv6 generally reduces the need for address sharing, but firewalls and privacy addresses still shape reachability and stability.

What an IP address can and cannot identify

A public address can indicate the network that announced it, its protocol version, and an approximate allocation geography. An IP lookup may associate it with a country, region, city estimate or time zone from a database. Those are inferences about network allocation, not GPS readings. Mobile routing, corporate gateways, satellite service, anycast, proxies and VPNs can move the visible endpoint far from the user.

An address alone usually cannot identify a named person. Multiple people may share one address, one person may use several, and reassignment changes the subscriber over time. Connecting an address to an account or subscriber requires additional records and a trustworthy timestamp. Even then, it may identify a connection rather than the individual behind an action.

Prefixes and subnet size

CIDR notation adds a slash and prefix length. In 192.0.2.0/24, the first 24 bits identify the network and eight bits remain, yielding 256 total values. Longer prefixes cover smaller ranges. IPv4 network and broadcast conventions affect traditional usable-host counts; /31 point-to-point networks and /32 single-host routes are special cases. IPv6 has no broadcast address, and a common /64 contains an enormous mathematical range. The Subnet / CIDR Calculator shows boundaries and handles those semantics explicitly.

A result is a snapshot

The visible address depends on the path taken at the time of the request. Switching from Wi-Fi to mobile data, enabling a VPN, using an IPv4-only or IPv6-capable destination, or passing through a corporate proxy can change it. Treat an IP result as useful network evidence with a time and context—not as a universal identity.