UUID Generator

What is a UUID?

UUID stands for Universally Unique Identifier. It is a 128-bit value written as 32 hexadecimal digits in five hyphen-separated groups of 8-4-4-4-12, for example 550e8400-e29b-41d4-a716-446655440000. The current specification is RFC 9562, published in 2024, which obsoletes the older RFC 4122.

The point of a UUID is that any machine can create one on its own and still expect it to be unique across every other machine, with no central registry, no database sequence and no network round trip. That property is what makes UUIDs the default identifier in distributed systems.

How does a UUID work?

Uniqueness comes from the size of the space rather than from bookkeeping. A version 4 UUID draws 122 random bits, which is about 5.3 × 10³⁶ possible values. Version 1 and version 7 instead combine a timestamp with random or sequential bits, so two UUIDs produced at different moments differ by construction, and two produced in the same moment differ by their random tail.

Six bits are always spoken for: four for the version, in the 13th hexadecimal digit, and two or three for the variant, at the start of the 17th digit. That is why the 17th character of a modern UUID is always 8, 9, a or b.

UUID vs GUID

GUID, short for Globally Unique Identifier, is the name Microsoft uses for the same 128-bit identifier family. For everyday purposes a GUID is a UUID: the same 32 hexadecimal digits, the same 8-4-4-4-12 grouping, the same version and variant fields.

The differences are conventions rather than substance:

• Microsoft tooling usually prints GUIDs wrapped in braces, as {550e8400-…}, and often in uppercase.
• The classic Windows binary layout stores the first three fields in little-endian order, so the same value can appear byte-swapped when read from a raw buffer.
• Some Microsoft-generated values use variant 2, the legacy Microsoft variant, rather than the RFC variant.

So a GUID and a UUID are interchangeable in text form, but do not assume every GUID implementation matches RFC 9562 in every detail. Use the Braces wrapper above to produce GUID-style output.

Are UUIDs truly unique?

Not guaranteed — designed so that a collision is negligibly unlikely. For random v4 UUIDs, reaching a 50% chance of even one collision takes roughly 2.7 × 10¹⁸ values, which is about 86,000 years of generating a million UUIDs every second. Uniqueness also depends on the source of randomness: a weak generator can produce collisions far sooner, which is why this tool refuses to fall back to Math.random().

How to generate UUIDs

On this page:

• Choose a version — v4 unless you have a reason to prefer time ordering.
• Set the quantity, from 1 to 1000, or tap a preset.
• Pick the letter case, hyphens and optional wrapper.
• Press Generate UUIDs, or Ctrl/⌘ + Enter.
• Copy a single row, copy the whole list, or download TXT or CSV.

In code:

• JavaScript — crypto.randomUUID()
• Python — uuid.uuid4()
• Java — java.util.UUID.randomUUID()
• Go — uuid.New() (google/uuid)
• C# — Guid.NewGuid()
• PostgreSQL — gen_random_uuid()
• Linux shell — uuidgen

How to validate a UUID

Length alone proves nothing. A correct check confirms three things: exactly 32 hexadecimal digits, the hyphens sitting after the 8th, 13th, 18th and 23rd characters, and — if you care about the version — that the variant digit marks it as an RFC 9562 value.

A canonical-form regular expression looks like this:

^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$

Normalise before you compare. Strip any {} or urn:uuid: wrapper, lowercase the digits and restore the hyphens, then compare the canonical strings — never the raw input. The validator above does exactly that and offers a one-click Normalize action.

Reading the version yourself

The version is the 13th hexadecimal digit — the first character of the third group. In 550e8400-e29b-41d4-a716-446655440000 that is 4. The variant lives at the start of the 17th digit, the first character of the fourth group.