ULID Generator

What is a ULID? (Universally Unique Lexicographically Sortable Identifier)

A ULID is a modern 128-bit identifier designed to resolve the limitations of traditional UUIDs while remaining 100% binary-compatible with UUID storage. Proposed by Alizain Feerasta in 2016, ULID combines the time-ordered sorting of auto-incrementing integers with the decentralized generation of cryptographic UUIDs.

Unlike standard UUIDs that require 36 characters with hyphens (e.g. f47ac10b-58cc-4372-a567-0e02b2c3d479), a ULID is encoded into 26 alphanumeric characters using Crockford's Base32 (e.g. 01ARZ3NDEKTSV4RRFFQ69G5FAV), making it compact, case-insensitive, and clean for HTTP URLs and REST APIs.

Anatomy & Bit Layout of a ULID

A ULID consists of two distinct components totaling 128 bits (16 bytes):

Why Crockford's Base32?

Crockford's Base32 alphabet uses characters 0123456789ABCDEFGHJKMNPQRSTVWXYZ. This encoding was chosen specifically over Hexadecimal or Base64 for three critical reasons:

• Human Error Prevention: Excludes letters I, L, O, and U to eliminate reading ambiguity with numbers 1 and 0.
• Accidental Profanity Defense: Excluding letter U prevents the unintentional generation of offensive words in public-facing identifiers.
• Case Insensitivity: Both uppercase (01ARZ...) and lowercase (01arz...) parse to identical binary values, preventing database collation bugs.

ULID vs. UUID v4 vs. UUID v7: Feature Comparison

The table below summarizes the key trade-offs between ULID, legacy random UUIDv4, and the modern IETF RFC 9562 UUIDv7:

Sub-Millisecond Monotonicity

When high-throughput systems generate multiple ULIDs within the same millisecond, a standard random generator might yield non-sequential values. Monotonic ULID implementations detect multiple calls within the identical timestamp and increment the 80-bit random component by 1. This ensures that every newly minted ULID remains strictly greater than its predecessor.

Production Code Implementation Library

// Node.js / TypeScript: npm install ulidx
import { ulid, monotonicFactory } from 'ulidx';

// Standard ULID
const id = ulid();
console.log(id); // e.g. "01ARZ3NDEKTSV4RRFFQ69G5FAV"

// Monotonic ULID for guaranteed sorting in the same millisecond
const monotonicUlid = monotonicFactory();
const id2 = monotonicUlid();

# Python: pip install python-ulid
import ulid

# Generate a new ULID
id = ulid.new()
print(id.str)       # 01ARZ3NDEKTSV4RRFFQ69G5FAV
print(id.timestamp) # Extracted epoch timestamp
print(id.datetime)  # Extracted datetime object

// Go: go get github.com/oklog/ulid/v2
package main

import (
    "fmt"
    "math/rand"
    "time"
    "github.com/oklog/ulid/v2"
)

func main() {
    t := time.Now()
    entropy := ulid.Monotonic(rand.New(rand.NewSource(t.UnixNano())), 0)
    id := ulid.MustNew(ulid.Timestamp(t), entropy)
    fmt.Println(id.String())
}

// Cargo.toml: ulid = "1.1"
use ulid::Ulid;

fn main() {
    let id = Ulid::new();
    println!("{}", id); // e.g. "01ARZ3NDEKTSV4RRFFQ69G5FAV"
    
    // Extract timestamp
    println!("Epoch ms: {}", id.timestamp_ms());
}