UUID Generator

How to generate UUIDs

1. Type how many identifiers you need in the "How many" box. The tool accepts 1 to 500; a larger number is capped at 500 and an empty or zero value falls back to 1.
2. Choose the format you want: Uppercase, No hyphens and Braces can be combined freely. The list is regenerated whenever you change a box, so you do not have to press anything.
3. Press Generate for a fresh batch with the same settings, then Copy all to put the whole list on the clipboard, one UUID per line.

Note that every change of an option creates brand-new values rather than re-formatting the old ones. If you need the same identifier in two formats, copy it first and convert it in your editor.

What is inside a version 4 UUID

A UUID is 128 bits, shown as 32 hexadecimal digits grouped 8-4-4-4-12. In a version 4 UUID most of those bits are random, but two fields are fixed so that software can recognise the type. The first digit of the third group is always 4 (the version), and the first digit of the fourth group is always one of 8, 9, a or b (the variant). That leaves 122 random bits.

Take f47ac10b-58cc-4372-a567-0e02b2c3d479. The third group starts with 4 and the fourth with a, so it is a valid v4 UUID. If you ever see a different digit in those positions, the value came from some other version or generator.

This page asks the browser for crypto.randomUUID() when it exists. In older browsers it falls back to filling 16 bytes with crypto.getRandomValues and then setting the version and variant bits by hand, which produces the same layout. Nothing is sent anywhere, and no list of earlier results is kept.

How likely is a duplicate?

With 122 random bits there are about 5.3 × 10³⁶ possible values. Because of the birthday effect, you would need to generate on the order of 2⁶¹ (roughly 2.3 × 10¹⁸) UUIDs before the chance of any two matching reaches one half. For a table of a few million rows the probability is far smaller than the chance of a hardware fault, so most systems simply treat v4 UUIDs as unique. A database unique constraint is still a sensible safety net, because it costs nothing and catches bugs such as copying the same value twice.

Common uses

• Primary keys that can be created on the client or on several servers without asking a central counter for the next number.
• Test data and fixtures: paste 50 identifiers into a seed file or a spreadsheet column.
• Correlation or request IDs attached to log lines so one user action can be traced across services.
• Unique file names for uploads, which avoids collisions and does not reveal the original name.
• Windows and .NET work: the Braces option gives the {...} form that registry files, installer projects and some configuration files expect; they usually call the value a GUID.

Choosing a format

The standard textual form is lowercase with hyphens, and that is what most libraries print and parse. Uppercase is the same number, just written differently; comparisons should be case-insensitive or you should normalise to one case before storing. The no-hyphen form is 32 characters and is handy for URLs, file names or systems that dislike dashes, but check that the receiving code accepts it, because strict parsers sometimes demand the hyphens.

Tips and limits

• A UUID is an identifier, not a secret. Do not use one as a password or as the only protection of a private link; use a purpose-built token generator for that.
• Random UUIDs have no order. In some databases, inserting random keys into an indexed column scatters writes across the index and can be slower than sequential keys. Newer time-ordered UUID versions exist for that case; this tool only makes version 4.
• Version 4 contains no timestamp, no machine address and no name, so a UUID reveals nothing about when or where it was created.
• The output box is read-only. Edit the copied text elsewhere if you need to tweak it.

More questions