ULID Generator Free Online Tool 2026

Last updated March 22, 2026

What Is a ULID?

A ULID (Universally Unique Lexicographically Sortable Identifier) is a 26-character string that combines a millisecond-precision timestamp with random bits. Unlike UUID v4, ULIDs sort chronologically, making them ideal as database primary keys in distributed systems.

ULID Format Breakdown

A ULID looks like this:

01ARZ3NDEKTSV4RRFFQ69G5FAV

• Characters 1–10: 48-bit timestamp (milliseconds since Unix epoch)
• Characters 11–26: 80 bits of cryptographically secure randomness
• Total: 128 bits — same size as a UUID
• Encoding: Crockford's Base32 (no I, L, O, U to avoid ambiguity)

Generating a Single ULID

Use the generator to produce one ULID at a time. Each click produces a new, unique value:

01HN2K5QXYZ3NDEKTSV4RRFFQ6
01HN2K5QXYZ3NDEKTSV4RRFFQ7
01HN2K5QXYZ3NDEKTSV4RRFFQ8

Notice the first 10 characters are identical — these three were generated in the same millisecond. The last 16 characters differ, ensuring uniqueness.

Bulk ULID Generation

Generate 10 ULIDs at once for seeding a test database:

01HN2K5QXYZ3NDEKTSV4RRFFQ6
01HN2K5QXYZ3NDEKTSV4RRFFQ7
01HN2K5QXYZ3NDEKTSV4RRFFQ8
01HN2K5QXYZ3NDEKTSV4RRFFQ9
01HN2K5QXYZ3NDEKTSV4RRFFQA
01HN2K5QXYZ3NDEKTSV4RRFFQB
01HN2K5QXYZ3NDEKTSV4RRFFQC
01HN2K5QXYZ3NDEKTSV4RRFFQD
01HN2K5QXYZ3NDEKTSV4RRFFQE
01HN2K5QXYZ3NDEKTSV4RRFFQF

These are already in chronological order — no ORDER BY timestamp needed.

Decoding a ULID

Paste an existing ULID to extract its timestamp:

Input:  01HN2K5QXYZ3NDEKTSV4RRFFQ6

Decoded:
  Timestamp component: 01HN2K5QXY
  Unix timestamp (ms): 1705312847123
  Human-readable time: 2024-01-15 08:20:47.123 UTC

  Random component:    Z3NDEKTSV4RRFFQ6
  Random bits:         80 bits of entropy

Using ULIDs as Database Primary Keys

Replace UUID v4 with ULID in a PostgreSQL table:

-- UUID v4 (random, no natural sort order)
CREATE TABLE orders_old (
  id UUID DEFAULT gen_random_uuid() PRIMARY KEY,
  customer_id INT,
  created_at TIMESTAMPTZ DEFAULT NOW()
);

-- ULID (sortable, no separate timestamp needed for ordering)
CREATE TABLE orders (
  id CHAR(26) PRIMARY KEY,  -- store as CHAR(26)
  customer_id INT
);

-- Insert with ULID
INSERT INTO orders (id, customer_id)
VALUES ('01HN2K5QXYZ3NDEKTSV4RRFFQ6', 42);

-- Retrieve in creation order using just the ID
SELECT * FROM orders ORDER BY id;

ULID in JavaScript

import { ulid } from 'ulid';

// Generate a new ULID
const id = ulid();
console.log(id); // 01HN2K5QXYZ3NDEKTSV4RRFFQ6

// Generate with a specific timestamp (for testing)
const fixedId = ulid(1705312847123);

// Decode timestamp from existing ULID
import { decodeTime } from 'ulid';
const timestamp = decodeTime('01HN2K5QXYZ3NDEKTSV4RRFFQ6');
console.log(new Date(timestamp)); // 2024-01-15T08:20:47.123Z

ULID in Python

import ulid

# Generate a new ULID
u = ulid.new()
print(str(u))  # 01HN2K5QXYZ3NDEKTSV4RRFFQ6

# Extract timestamp
print(u.timestamp().datetime)  # 2024-01-15 08:20:47.123000+00:00

# Parse an existing ULID string
parsed = ulid.from_str('01HN2K5QXYZ3NDEKTSV4RRFFQ6')
print(parsed.timestamp().int)  # 1705312847123

ULID vs UUID Comparison

Format         Example                                  Sortable  Size
-----------    ---------------------------------------  --------  ----
UUID v4        550e8400-e29b-41d4-a716-446655440000     No        128 bit
UUID v1        6ba7b810-9dad-11d1-80b4-00c04fd430c8     Yes*      128 bit
UUID v7        018e3a5f-1234-7abc-8def-000000000001     Yes       128 bit
ULID           01HN2K5QXYZ3NDEKTSV4RRFFQ6              Yes       128 bit
Sequential INT 1, 2, 3, 4 ...                           Yes       32/64 bit

* UUID v1 is sortable but exposes MAC address

ULID for Distributed Event Logging

In a microservices architecture, multiple services generate events simultaneously. Using ULIDs as event IDs allows merging logs from different services in chronological order:

Service A event: 01HN2K5QXYZ3NDEKTSV4RRFFQ6  (08:20:47.123)
Service B event: 01HN2K5QXYZ3NDEKTSV4RRFFQ9  (08:20:47.123)
Service C event: 01HN2K5QXYZ3NDEKTSV4RRFFQZ  (08:20:47.124)

-- Merge and sort across services
SELECT service, event_id, payload
FROM events
ORDER BY event_id;  -- chronological order, no timestamp column needed

B-Tree Index Performance

ULIDs improve write performance on indexed tables compared to UUID v4:

• UUID v4: random inserts scatter across the entire B-tree, causing page splits and fragmentation
• ULID: new records always append near the end of the index (monotonically increasing), minimizing fragmentation
• Result: significantly faster INSERT performance on high-write tables

-- Benchmark: 1 million inserts
-- UUID v4 primary key: ~45 seconds, high fragmentation
-- ULID primary key:    ~12 seconds, minimal fragmentation

Common Use Cases

• Database primary keys where chronological ordering matters
• API resource identifiers (sortable by creation time)
• Distributed event IDs across microservices
• File names that sort by creation date
• Cache keys with built-in time ordering
• Audit log entry identifiers