UUID / nanoid
Generate IDs in bulk with options
Generate cryptographically random UUID v4 or nanoid identifiers in bulk, with options for uppercase and hyphen formatting, computed locally using the browser's `crypto.getRandomValues` — no server call, no rate limit, no possibility of a service logging or reusing the IDs it gives you. A UUID (Universally Unique Identifier) is a 128-bit value, conventionally written as 32 hex digits in 8-4-4-4-12 groups (e.g. `550e8400-e29b-41d4-a716-446655440000`), chosen so randomly that two independently generated UUIDs are vanishingly unlikely to ever collide — making them the standard choice for database primary keys, distributed-system request IDs, and session tokens where coordinating a central counter isn't practical. nanoid is a newer, more compact alternative (typically 21 URL-safe characters) that's shorter to store and display while keeping comparable collision resistance.
What's the difference between UUID v4 and other UUID versions?
UUID v4 is generated from random or pseudo-random numbers and is the most common choice today because it requires no coordination or shared state. Other versions encode different information: v1 embeds a timestamp and the generating machine's MAC address (which can leak information), v3/v5 are deterministic hashes of a namespace and name (useful when you need the same input to always produce the same ID), and v7 (newer) embeds a sortable timestamp prefix for database-friendly ordering.
How likely is a UUID v4 collision?
With 122 random bits, you'd need to generate roughly 2.7 quintillion UUIDs before a 50% chance of any collision — in practice, for any real-world application's lifetime, the probability is treated as zero.
Why use nanoid instead of UUID?
nanoid produces shorter IDs (21 characters vs UUID's 36) using a larger, URL-safe alphabet, which makes IDs more compact in URLs, logs, and database indexes while keeping a comparable collision rate to UUID v4 for typical application use.