UUID Generator
Click Generate and get a UUID v4, or generate up to 100 at once for seeding test data. Uses the Web Crypto API so the randomness is cryptographically strong. Nothing is sent anywhere.
••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••Generate UUID v4 (Universally Unique Identifiers) for your projects. UUIDs are 128-bit numbers used to uniquely identify information in computer systems. All processing happens locally in your browser - no data is sent to any server.
How to generate UUIDs
Three steps. Generating 50 at once for a seed file takes the same time as generating one.
- 01
Set the count
1 for a single UUID, up to 100 for a batch. Useful when seeding a database or generating test fixtures. Paste the whole list at once.
- 02
Click Generate
Each click produces a fresh set of cryptographically random UUID v4 values using the Web Crypto API. No two will ever be the same.
- 03
Copy what you need
Copy individual UUIDs or the entire list at once. Paste them into your code, migration file, or config.
Where UUIDs actually get used
Primary keys that don't leak record counts
Auto-increment integers tell anyone who looks at a URL how many records you have. UUIDs don't. They also survive merges between databases without collision.
Distributed systems without a central ID server
Multiple services can generate IDs independently and they'll never clash. No coordination needed, no single point of failure.
Seeding test databases
Generate a batch of 20 or 50 UUIDs and paste them directly into your fixture files. Realistic-looking IDs, no manual typing.
File uploads that don't expose sequence
Name uploaded files with UUIDs instead of sequential numbers. Users can't guess other filenames, and collisions are impossible.
Session and token IDs
UUIDs work well as session identifiers and one-time link tokens, long enough to be unguessable, short enough to fit in a cookie or header.
REST and GraphQL resource IDs
Expose UUIDs in your API instead of database integer IDs. Clients can't enumerate records by incrementing a number.
How unlikely a duplicate really is
A version 4 UUID has 128 bits, of which 6 are fixed to mark the version and variant. The other 122 are random, which gives 2¹²² possible values, about 5.3 × 10³⁶. Each UUID here comes fromcrypto.randomUUID(), the browser's cryptographic random source.
Collisions follow the birthday problem, so the chance of any two matching grows with the square of how many you make. Even so, you'd need to generate around 103 trillion UUIDs before the chance of a single duplicate reached one in a billion. For any real application, the risk of two random v4 UUIDs clashing is far smaller than the risk of a bug in the code that stores them.
Version 4 or version 7?
RFC 9562, published in 2024, added version 7, which starts with a millisecond timestamp and fills the rest with random bits. Because v7 values sort in creation order, they keep database indexes compact, while random v4 values land all over a B-tree index. This tool generates v4 only. Use it for tokens, file names and IDs where order doesn't matter, and consider your database or language library's v7 support for high-volume primary keys.