UUID Guide — What They Are, How to Generate Them, and When to Use Them

A UUID (Universally Unique Identifier) is a 128-bit identifier formatted as a 32-character hexadecimal string separated into five groups:

550e8400-e29b-41d4-a716-446655440000

They're designed to be generated independently across distributed systems without any central coordination — and still be unique. The probability of generating a duplicate UUID v4 is astronomically small (around 1 in 5.3 × 10³⁶).

Use Intoolhub's UUID Generator to generate UUIDs in any version.

UUID Versions Explained

There are five official UUID versions, each using a different generation strategy:

UUID v1 — Timestamp + MAC address

Combines the current timestamp with the generating machine's MAC address. This makes v1 UUIDs time-ordered and potentially traceable back to a specific machine — which is a privacy concern in public-facing systems.

Use when: You need sortable IDs in a trusted internal environment.

UUID v4 — Random

The most widely used version. 122 bits of cryptographically random data. No timestamp, no MAC address — just entropy.

f47ac10b-58cc-4372-a567-0e02b2c3d479

Use when: You need a unique identifier and don't need sortability. The right default for most applications.

UUID v5 — Namespace + SHA-1 hash

Deterministic: given the same namespace and name, it always produces the same UUID. Uses SHA-1 internally.

// Always produces the same UUID for the same inputs
uuidv5('https://example.com', uuidv5.URL)
// → 'c74a196f-f19d-5ea9-bffd-a2742199079c'

Use when: You need a repeatable, collision-resistant ID for a known value (e.g., turning a URL or username into a UUID).

UUID v3 — Namespace + MD5 hash

Same concept as v5 but uses MD5 instead of SHA-1. MD5 is considered cryptographically broken — prefer v5 for new systems.

UUID v7 (Draft RFC)

A modern, time-ordered version designed to replace v1 without its privacy issues. Embeds a Unix millisecond timestamp in the first 48 bits, making UUIDs naturally sortable as database primary keys.

018e57d2-2c00-7000-8000-000000000000
└──────────────┘
 ms timestamp (sortable)

Use when: You need database-friendly, sortable UUIDs without exposing MAC addresses.

UUID Structure

Every UUID follows the same format:

xxxxxxxx-xxxx-Mxxx-Nxxx-xxxxxxxxxxxx

• M — version digit (1, 3, 4, 5, or 7)
• N — variant digit (always 8, 9, a, or b for RFC 4122 UUIDs)

Generating UUIDs in Code

JavaScript / TypeScript

// Node.js 14.17+ and modern browsers
import { randomUUID } from 'crypto'
const id = randomUUID() // v4

// Or use the uuid package for more control
import { v4 as uuidv4, v5 as uuidv5 } from 'uuid'
const id = uuidv4()

Python

import uuid

uuid.uuid4()  # v4 random
uuid.uuid5(uuid.NAMESPACE_URL, 'https://example.com')  # v5

PostgreSQL

-- Enable the extension
CREATE EXTENSION IF NOT EXISTS "uuid-ossp";

-- Generate
SELECT uuid_generate_v4();

-- Or with pgcrypto (v4 only)
SELECT gen_random_uuid();

MySQL

SELECT UUID();  -- generates v1

UUIDs as Database Primary Keys

The classic advice is to use sequential integers (AUTO_INCREMENT, SERIAL) as primary keys because they perform better for B-tree indexes. UUIDs are random, causing index fragmentation.

UUID v7 solves this by being time-ordered, giving you near-sequential inserts without the coordination overhead of a centralized counter.

If you must use v4 UUIDs as primary keys in high-insert databases, consider:

• Storing as BINARY(16) not VARCHAR(36) (saves 20 bytes per row)
• Using ULIDs as an alternative (also sortable, slightly more compact)

UUIDs vs ULIDs vs NanoIDs

For most new applications, UUID v7 or ULID is the better choice over UUID v4 when using IDs as database primary keys.

→ Open UUID Generator