UUID Decoder

A UUID is a 128-bit number, usually written as 32 hexadecimal digits in five groups, that systems use as an identifier without asking a central authority. Because the number of possible values is enormous, two systems can generate IDs independently and almost never collide. Not every UUID is a random string, though. Several versions embed a time, a name or a machine address, and knowing which one you hold tells you what an ID reveals. This decoder shows the version, variant and fields of any UUID.

The version and the variant

Two small fields describe the layout. The version is the first hex digit of the third group (the 13th digit overall): 1 for time-based, 3 and 5 for name-based, 4 for random, 6 and 7 for the newer time-ordered kinds and 8 for custom. The variant is carried by the top bits of the fourth group, and for ordinary UUIDs it reads 10xx, which the page labels as RFC 4122 / RFC 9562. Anything that does not follow it, such as the old Microsoft layout, is reported differently.

A worked example

Press Sample to decode six strings. c232ab00-9414-11ec-b3c8-9f6bdeced846 is version 1, time-based: timestamp 2022-02-22T19:22:22.000Z, which is 138648505420000000 ticks of 100 nanoseconds since 1582-10-15, clock sequence 13256 and node 9f:6b:de:ce:d8:46. 1ec9414c-232a-6b00-b3c8-9f6bdeced846 is the version 6 form of the same moment, with the same clock sequence and node. 017f22e2-79b0-7cc3-98c4-dc0c0c07398f is version 7: Unix milliseconds 1645557742000, which is the same instant. f47ac10b-58cc-4372-a567-0e02b2c3d479 is version 4, random, with no field to decode. 00000000-0000-0000-0000-000000000000 is the nil UUID, and not-a-uuid is reported as invalid.

Timestamps and sorting

Version 1 splits its timestamp into three pieces and stores the low part first, so values do not sort by time as text. Version 6 reorders the same bits, high part first, and version 7 uses the simpler 48-bit Unix millisecond count followed by random bits. For database keys that grow with time, 6 and 7 behave like ordered identifiers, while 4 behaves like random noise.

What a UUID reveals

A version 1 UUID written by a computer with a network card carries that card’s MAC address in the node field and the exact time of creation. Together they identify the machine and when it acted, which has been used to trace documents to their authors. Many libraries use a random node instead, which the multicast bit in the node’s first byte indicates. Versions 4 and 8 reveal nothing by design, and version 7 shows only the time.

Name-based UUIDs

Versions 3 and 5 hash a namespace and a name, with MD5 and SHA-1 respectively, so the same name always yields the same UUID. They are useful for stable identifiers derived from URLs or domain names. The page identifies them but cannot reverse the hash.

Limits

Decoding reads the fields, which are only as honest as the generator that wrote them. The page does not test for uniqueness or ownership. Everything runs in the browser.

Related tools

• UUID Generator creates new values of the common versions.
• Timestamp Converter converts the Unix times found in version 7 values.
• Hash Identifier identifies hash strings, which look similar at first glance.
• Base64 Encoder converts an ID to a shorter text form.