A UUID (Universally Unique Identifier) is a 128-bit label used to uniquely identify objects in computer systems. The standard format is 8-4-4-4-12 hexadecimal digits separated by hyphens, e.g. 550e8400-e29b-41d4-a716-446655440000.

UUID v4 is randomly generated — 122 bits of random data plus 6 bits of version/variant information. It's the most commonly used UUID version for general-purpose unique IDs in databases and APIs.

Are UUIDs truly unique?

UUID v4 has 2^122 possible values (~5.3 × 10^36). The probability of generating two identical UUIDs is astronomically small — you'd need to generate 1 billion UUIDs per second for 85 years to have a 50% chance of a single collision.

What's the difference between UUID and GUID?

GUID (Globally Unique Identifier) is Microsoft's term for the same concept. UUIDs and GUIDs follow the same RFC 4122 specification and are interchangeable.

UUID Generator

Generate UUID v4 instantly — bulk, formatted, cryptographically random

Count Format

Generate UUIDs in Code

// Browser (Web Crypto API)
const uuid = crypto.randomUUID(); // native, no library needed

// Node.js 14.17+
const { randomUUID } = require('crypto');
const uuid = randomUUID();

// npm: uuid package
import { v4 as uuidv4 } from 'uuid';
const id = uuidv4();

import uuid

# Generate a random UUID v4
uid = uuid.uuid4()
print(uid)           # 550e8400-e29b-41d4-a716-446655440000
print(str(uid))      # as string
print(uid.hex)       # no hyphens: 550e8400e29b41d4a716446655440000

import "github.com/google/uuid"

// go get github.com/google/uuid
id := uuid.New()
fmt.Println(id.String())

// Standard library (Go 1.20+)
import "crypto/rand"
// manual implementation or use google/uuid

-- PostgreSQL
SELECT gen_random_uuid();           -- built-in v4
SELECT uuid_generate_v4();          -- uuid-ossp extension

-- MySQL 8+
SELECT UUID();                      -- returns v1, not v4
SELECT REPLACE(UUID(), '-', '');   -- no hyphens

-- SQL Server
SELECT NEWID();                     -- equivalent to UUID v4

UUID Versions

v1 — Time-based

Uses timestamp + MAC address. Sortable chronologically but leaks machine info. Used in Cassandra, Snowflake-style IDs.

v4 — Random ★

122 bits of random data. Most common for general-purpose IDs in databases, APIs, and distributed systems.

v5 — Name-based

SHA-1 hash of a namespace + name. Same input always produces same UUID. Useful for deterministic IDs.

v7 — New standard

Time-ordered random UUID (RFC 9562). Sortable like v1 but random like v4. Growing adoption in modern databases.

Frequently Asked Questions

A UUID (Universally Unique Identifier) is a 128-bit label in the format 8-4-4-4-12 hex digits. Example: 550e8400-e29b-41d4-a716-446655440000. Used to uniquely identify records in databases, APIs, and distributed systems.

Practically yes. UUID v4 has 2^122 ≈ 5.3 × 10^36 possible values. Generating 1 billion UUIDs/second for 85 years gives only a 50% chance of a single collision — effectively impossible in practice.

GUID (Globally Unique Identifier) is Microsoft's name for the same concept. Both follow RFC 4122 and are completely interchangeable. .NET uses GUID, most other platforms use UUID.

Yes, but v4 UUIDs are random and cause index fragmentation in B-tree indexes. For high-insert databases, prefer UUID v7 (time-ordered) or ULID which sort chronologically and are more index-friendly.

Related Tools

Hash

Hash Generator

MD5, SHA-256, SHA-512

Encode

Base64 Encoder

Encode text and files

Generate

QR Generator

QR codes for URLs, WiFi

Convert

Epoch Converter

Unix timestamp ↔ date

Prefer UUIDv7 for databases

In 2026, if you’re choosing a new UUID format for database keys, prefer UUIDv7 over random UUIDv4 when your stack supports it. UUIDv7 keeps the same 128-bit size but starts with a timestamp, so new rows insert in near-sorted order instead of scattering across indexes. That usually means smaller index fragmentation and better write performance in PostgreSQL, MySQL, and SQLite. Common mistake: generating v4 everywhere by habit, then wondering why primary-key indexes bloat. Keep v4 for pure randomness needs; use v7 for ordered IDs.