UUID, ULID, and Unique Identifier Standards

Globally unique identifiers are fundamental building blocks of distributed software systems. When multiple services, databases, or clients need to create records independently without central coordination, unique ID schemes prevent collisions without a shared counter or sequence.

UUID v4 (Universally Unique Identifier)

UUID v4 is the most widely used format, standardised in RFC 4122. It consists of 128 random bits formatted as 32 hexadecimal digits in five groups separated by hyphens: xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx. The 4 denotes version 4 (random), and the y position encodes the variant. The probability of collision between two random v4 UUIDs is astronomically low — approximately 1 in 5.3 × 10³⁶.

ULID (Universally Unique Lexicographically Sortable Identifier)

ULIDs encode a 48-bit millisecond timestamp followed by 80 random bits, formatted as a 26-character Crockford Base32 string. Unlike UUIDs, ULIDs sort chronologically by creation time, making them ideal for database primary keys where insertion order matters for index performance.

Short IDs

For applications where UUIDs are too verbose — URL slugs, invite codes, session tokens — short random identifiers (8–12 alphanumeric characters) provide sufficient uniqueness for smaller datasets while remaining human-readable and URL-safe.

Why Generate IDs Client-Side?

UUID generation using crypto.getRandomValues() is cryptographically secure and requires no server round-trip. Generating IDs locally is faster, works offline, and ensures your ID schemes are never exposed to third-party logging.