UUID Generator

A version 4 UUID is 122 bits of randomness dressed up as 36 characters, and its whole value is that two machines can each mint one without talking to each other and never collide. The UUID Generator produces them in batches of up to 100, drawing every byte from the browser's cryptographic random source rather than a general-purpose one.

Where it can, generation goes through the native randomUUIDcall. Where that is unavailable, on an older browser or a page served over plain HTTP, it falls back to pulling 16 cryptographic random bytes and setting the version and variant bits by hand. Both paths are equally unpredictable, which matters because a UUID used as a token or an unguessable URL is only as private as the randomness behind it.

One identifier appears on load so the page is never empty. Each row reveals its own copy button on hover, the text is set to select all at once on a single click, and a Copy All button appears as soon as a batch holds more than one, joining them with newlines for pasting into a spreadsheet or a seed file.

What Each Part of a v4 UUID Means

• The layout is 8, 4, 4, 4 and 12 hex digits separated by hyphens, 32 digits and 36 characters in total.
• The first digit of the third group is always 4. That is the version marker, and it is how you tell a random UUID from a time-based version 1 at a glance.
• The first digit of the fourth group is always 8, 9, a or b. Those 2 fixed bits are the variant field, marking the identifier as following the standard layout.
• Those 6 fixed bits are why the entropy is 122 rather than 128. It remains an enormous number: generating a billion per second for a century leaves the chance of a single collision far below anything worth designing around.
• Set Batch Count anywhere from 1 to 100 and press Generate. The list is replaced rather than appended to, so the previous batch is gone once you press it again.

Not a Substitute for a Database Key Strategy

• Random UUIDs make poor clustered primary keys. Each insert lands at a random point in the index, which fragments B-tree pages and can cost a great deal of write throughput on a large table. Sequential variants such as UUIDv7 exist for exactly this reason.
• Stored as text a UUID costs 36 bytes against the 16 it needs as a native type, and that overhead is multiplied by every index that includes the column.
• Version 4 carries no timestamp, so you cannot sort by it or recover a creation order from it. Keep a separate created-at column.
• An identifier that is unguessable is still not a credential. Anyone holding the URL holds the access, and a UUID in a shared link ends up in browser history, referrer headers and proxy logs.
• These are generated in your browser and never recorded anywhere, which also means there is no way to check whether one you were given came from here.