What are IP addresses?
An IP address is a fixed-length value used by Internet Protocol for packet delivery within a routing scope. IPv4 addresses are 32 bits and IPv6 addresses are 128 bits. An address is associated with an interface or addressing function; it is not a permanent identity for a device, organization, or person.
Names, addresses, routes, registry records, and contractual authority describe different facts. Protocol history explains how the standards evolved, but it does not prove the present right to use, route, transfer, or lease a particular prefix. For current fundamentals, see the IPv4 address guide.
The early specifications before IPv4
Research specifications changed repeatedly while TCP and IP were being separated and deployed. The surviving primary-source record does not establish a simple series of globally standardized products called IPv1, IPv2, and IPv3. RFC 791 says its 1981 text was based on six earlier editions of the ARPA Internet Protocol specification and credits many contributors.
What does “IPv1” mean?
“IPv1” is a retrospective label sometimes applied to early experimental Internet Protocol work. It is not the name of an IETF Internet Standard or a globally deployed commercial protocol. Use dated specification identifiers when discussing an early design instead of treating “IPv1” as one finished product.
What happened to “IPv2” and “IPv3”?
These labels are also used informally for revisions made before version 4. The available standards do not support assigning a single feature list, release date, or worldwide deployment to products called IPv2 and IPv3. The defensible conclusion is narrower: several earlier specification editions preceded the version-4 documents.
When was IPv4 created?
RFC 760, published in January 1980, already described the version field as 4. RFC 791 replaced it in September 1981 and remains the defining early IPv4 specification. Calling 1981 the moment IPv4 was first created hides the earlier version-4 specification and the iterative work behind it.
- Address size: IPv4 uses 32-bit addresses, giving exactly 232, or 4,294,967,296, possible values before reservations and operating constraints.
- Delivery model: IP moves independent datagrams across interconnected networks; reliability, sequencing, and application behavior belong to other layers.
- Continued use: IPv4 remains widely operated because networks, applications, contracts, equipment, and address-management processes cannot all change at once.
When was IPv6 created?
RFC 1752, published in January 1995, recorded the IESG-accepted recommendation for what should replace the current Internet Protocol and used the name IPv6 for the selected IPng direction. RFC 1883 then specified Internet Protocol Version 6 in December 1995. RFC 2460 replaced that specification in December 1998. RFC 8200 replaced RFC 2460 in July 2017 and identifies IPv6 as Internet Standard STD 86. These are recommendation and specification milestones, not single dates of invention, commercial release, or worldwide deployment.
- Address size: IPv6 uses a finite 128-bit address space. It is vastly larger than IPv4, but not infinite.
- Security: IPv6 can use IPsec, but the protocol does not automatically authenticate or encrypt every packet. Current node requirements say implementations should support the IPsec architecture; deployment and key management still determine protection.
- Performance: IPv6 changes the header and removes router fragmentation, but it does not guarantee that every application or route will be faster. Measure the actual path, DNS, transport, filtering, and transition mechanisms.
Why is there no general-purpose IPv5 successor?
The value 5 was used for the experimental Internet Stream Protocol Version 2, or ST-II. RFC 1190, published in October 1990, says ST operated at the same layer as IP, maintained stream state, and used IP version number 5. ST-II supported experiments involving real-time streams; it was not the general-purpose successor to IPv4 later standardized as IPv6.
Key specification milestones
- January 1980: RFC 760 specifies Internet Protocol with version field value 4.
- September 1981: RFC 791 replaces RFC 760 and revises the IPv4 specification.
- October 1990: RFC 1190 documents experimental ST-II using version field value 5.
- January 1995: RFC 1752 records the IETF’s IPng recommendation and the IPv6 name.
- December 1995: RFC 1883 specifies Internet Protocol Version 6.
- December 1998: RFC 2460 replaces RFC 1883.
- July 2017: RFC 8200 replaces RFC 2460 and becomes Internet Standard STD 86.
What this history means today
IPv4 and IPv6 now coexist. A migration plan may use native IPv6, dual stack, translation, tunnelling, or a staged combination, depending on applications and network boundaries. The version number alone does not establish reachability, security, performance, or current authority over an address block.
For an operating IPv4 prefix, verify the exact CIDR, registry and contract records, authorized route origin, LOA, IRR and RPKI state, reputation evidence, abuse handling, term, renewal, and return conditions.
Apply protocol history to current IPv4 operations
Use the standards record to understand the protocol, then use current evidence to make a resource decision. Historic specifications do not substitute for live registry, routing, security, or contractual checks.



