DNSfinder.net
Transition Formats

IPv4 to IPv6 Converter

IPv4 and IPv6 are separate address spaces, so an IPv4 address has no single IPv6 equivalent. What it does have is a set of standard IPv6 spellings used by dual-stack software and transition mechanisms. Enter an IPv4 address and this converter shows all of them: the IPv4-mapped form, the NAT64 well-known-prefix form, the 6to4 prefix and the deprecated IPv4-compatible form, each with a copy button. It runs in your browser and sends nothing to a server.

Loading IPv4 to IPv6 Converter...

Querying live edge resolvers and network infrastructure...

How to use the IPv4 to IPv6 Converter

  1. Type an IPv4 address such as 192.0.2.33: four numbers between 0 and 255, no leading zeros.
  2. Read the IPv6 representations card, which lists each form with a short note on where it is used.
  3. Copy the IPv4-mapped form for log filters and socket code, or the NAT64 form to reach the host from an IPv6-only network.
  4. Take the info box seriously: none of these is the host's real IPv6 address, which only its AAAA record can tell you.

What to look for

IPv4-mapped, three ways

The card shows ::ffff:192.0.2.33, the same value with the last 32 bits in hex (::ffff:c000:221) and the full eight-group form 0000:0000:0000:0000:0000:ffff:c000:0221. All three are one address in the ::ffff:0:0/96 block, which is how a dual-stack socket presents an IPv4 peer to the application.

NAT64 and the well-known prefix

64:ff9b::192.0.2.33 is the address an IPv6-only client uses to reach the IPv4 host through a NAT64 translator running the well-known prefix from RFC 6052. The hex row shows the same address as 64:ff9b::c000:221, which is how it tends to appear in logs and routing tables.

6to4 and IPv4-compatible

2002:c000:221::/48 is the 6to4 prefix derived from the address: the 32 bits of the IPv4 address follow 2002 and the owner can number subnets beneath it. ::192.0.2.33 is the IPv4-compatible form, deprecated by RFC 4291 and shown only so you can recognise it in old configurations.

Guide

What each IPv6 form of an IPv4 address is for

Why there is no direct conversion

An IPv4 address is 32 bits and an IPv6 address is 128 bits. Nothing in the protocols maps one to the other: a server with the IPv4 address 192.0.2.33 has an IPv6 address only if its network assigned one, and that address is published separately in an AAAA record. The forms on this page are conventions for carrying an IPv4 address inside an IPv6 one, each tied to a specific mechanism.

That is why the result card carries an info box. People sometimes paste the IPv4-mapped form into a DNS zone as an AAAA record, which breaks IPv6 clients: they try to connect to ::ffff:192.0.2.33 as if it were a routable IPv6 address, and it is not. DNS lookup with the AAAA type shows what a host's real IPv6 address is.

IPv4-mapped addresses in sockets and logs

When a program opens an IPv6 socket without setting the IPV6_V6ONLY option, the operating system also accepts IPv4 connections on it and reports each IPv4 peer as an IPv4-mapped address. That is why a web server listening on [::]:443 logs an IPv4 visitor as ::ffff:203.0.113.5. The address is a representation inside the host. On the wire the packet was plain IPv4.

The practical use of the mapped form is matching. A firewall rule, a log filter or an allow-list that must work for both families can be written against the ::ffff:0:0/96 block, and a script that normalises addresses can strip the prefix to recover the IPv4 value. The reverse tool, IPv6 to IPv4, does that extraction and recognises the other mechanisms as well.

Reaching IPv4 hosts from IPv6-only networks

Mobile carriers and some corporate networks run IPv6-only clients and reach the IPv4 internet through NAT64. A DNS64 resolver synthesises AAAA answers for IPv4-only names by placing the IPv4 address inside the prefix 64:ff9b::/96, and the translator strips it again on the way out. The NAT64 row gives you that synthesised address so you can test the path directly, for instance with a ping or a port check from an IPv6-only host.

The well-known prefix is a default. Many operators use a network-specific prefix instead, in which case the address on this page will not match what their resolver returns. If a test fails, ask which prefix the network uses before concluding that the translator is down.

6to4 and the deprecated forms

6to4 (RFC 3056) gave any host with a public IPv4 address a whole IPv6 /48 by placing the address after the 2002: prefix, so 192.0.2.33 owns 2002:c000:221::/48, and tunnelled the packets over IPv4 to public relays. The relays were unreliable and the mechanism is deprecated for general use. The row is here because 2002: addresses still appear in old DNS zones and routing tables, and the prefix tells you which IPv4 gateway they came from.

The IPv4-compatible form ::192.0.2.33 predates all of this. RFC 4291 deprecated it and no current mechanism uses it, so treat it as documentation only. Note that 6to4 needs a public IPv4 address to mean anything: a private address such as 10.1.2.3 produces a syntactically valid prefix that no relay would ever route.

Knowledge Base & FAQ

IPv4 to IPv6 Converter FAQs

Common questions about the ipv4 to ipv6 converter.

You can write any IPv4 address in the IPv6 forms shown here, but that does not give the host IPv6 connectivity. A real IPv6 address is assigned by the network and published in an AAAA record. The converter produces representations used by transition mechanisms and software, not an address you can route to natively.