Overview
Somewhere in nearly every database, API and distributed system, things need names that will never collide — not today, not on another server, not in ten years. The UUID is the industry's answer: a 128-bit identifier, standardised and random enough that two independently generated ones will not match in the lifetime of the universe. Version 4 is the random flavour, and this page mints them properly.
You control the batch and the format. Generate anywhere from one to fifty at a time, keep the classic hyphenated grouping or strip it to a compact 32-character string, switch to uppercase hex for systems that expect it, then copy the whole batch or download it as a plain text file for scripts and fixtures.
What exactly is a UUID?
A UUID (universally unique identifier) is a 128-bit number written as 32 hexadecimal digits, usually grouped 8-4-4-4-12 with hyphens: something like 550e8400-e29b-41d4-a716-446655440000. Version 4 means 122 of those bits are random; a few fixed bits mark the version (the ‘4’ you can see in the third group) and the variant, exactly as the standard demands.
Uniqueness comes from arithmetic, not coordination. With 122 random bits there are about 5×10³⁶ possible values — generating a duplicate by chance is so unlikely that systems simply assume it will not happen, which is why UUIDs work as identifiers without a central registry handing them out.
Features
- Batches of 1 to 50 — one identifier for a quick test, fifty for an import or seed data.
- True version-4 format: the version nibble reads 4 and the variant bits follow the standard exactly.
- Hyphens on or off — classic grouped form or a compact 32-character string.
- Uppercase option for systems that expect capital hex digits.
- Copy all in one click — the whole batch to your clipboard, one UUID per line.
- Download as .txt — save the batch as a plain text file for scripts and fixtures.
- Cryptographic randomness from your browser's secure source, not Math.random.
Benefits
- Correct by construction. Version and variant bits set per the standard — no almost-UUIDs.
- Bulk-friendly. Fifty identifiers in one click beats fifty round trips.
- Format-flexible. Hyphenated, compact or uppercase to match whatever consumes them.
- Private. Generated locally; nothing about your project leaks anywhere.
Who can use it
Developers seeding databases and writing tests, API designers minting resource IDs, data engineers generating surrogate keys for imports, and students learning how distributed systems name things without a central authority. Anyone who has ever needed ‘some unique string, right now’ — this is the canonical answer.
How to use
- Choose how many UUIDs you need, from 1 to 50.
- Toggle hyphens and uppercase to match the format your system expects.
- Press generate and inspect the batch.
- Copy all to the clipboard, or download the batch as a .txt file.
- Paste them into your database, fixtures, config or code — each is ready to use.
Frequently asked questions
Can two UUIDs ever be the same?
In theory, yes; in practice, no. Version 4 has 122 random bits, giving about 5×10³⁶ possibilities. You would need to generate around a billion UUIDs per second for a century before a duplicate became likely. Systems treat collision as impossible, and the maths backs them up.
What's the difference between the hyphenated and compact forms?
Nothing but presentation — the same 128 bits, with or without the four hyphens grouping them. Some databases, APIs and config formats accept both forms; others are picky about which one they want. The toggle exists because being fussy about UUID formatting is a time-honoured tradition in software, and this page would rather adapt than argue.
Why not just use Math.random()?
Because Math.random() is deterministic: seeded the same way, it replays the identical sequence, which makes it predictable in principle. That is fine for shuffling a playlist but wrong for identifiers that must be unguessable. This page draws from your browser's cryptographic entropy pool — the same source that secures web connections.
Should UUIDs be kept secret?
They are identifiers, not passwords — random enough to be unguessable in practice, but not a substitute for authentication. Treat them as public-ish names that happen to be hard to enumerate, and put real access control in front of anything sensitive.
Version 4 vs version 7 — which should I use?
Version 4 (random) is the universal default and what this page generates — supported everywhere, understood by everyone. Version 7 (time-ordered) is newer and sorts nicely in databases, but library support is still catching up. For most purposes today, v4 remains the safe, boring and correct choice.
Conclusion
When things need names that will never collide — database rows, API resources, test fixtures — reach for a proper version-4 UUID. Generate a batch, pick hyphenated, compact or uppercase to suit the consumer, and get back to the interesting problems.