ULID Generator and Decoder

Generate and decode Universally Unique Lexicographically Sortable Identifiers (ULID). Extract millisecond timestamps, parse Crockford Base32, and generate monotonic batches.

Live ULID Generator

26-character, URL-safe, lexicographically sortable identifier.

Decode ULID Identifier

Enter 26-Character ULID String

Please enter a valid 26-character Crockford Base32 encoded ULID string.

Batch ULID Generator

Generate multiple sequential IDs for seed scripts or database testing.

High-Performance Distributed Primary Keys with ULID

Universally Unique Lexicographically Sortable Identifiers (ULIDs) resolve the performance drawbacks of traditional random UUIDv4s in relational and document databases (PostgreSQL, MySQL B-Trees). By combining a 48-bit millisecond timestamp with 80 bits of cryptographic randomness in 26 characters of URL-safe Crockford's Base32, ULIDs prevent database index fragmentation, provide natural time sorting, and remain 100% compatible with 128-bit UUID binary storage.

Key Technical Features

✓

Instant ULID Generation

Generate single or batch (up to 50) cryptographically secure Crockford Base32 ULID identifiers.

✓

Millisecond Timestamp Extraction

Decodes the leading 10 characters into human-readable ISO-8601 UTC dates, Unix epoch timestamps, and relative time.

✓

Monotonic Factory Support

Ensures sequential ordering even when multiple IDs are generated within the exact same millisecond.

✓

UUIDv4 Hex Cross-Conversion

Convert ULIDs to standard 8-4-4-4-12 hex UUID format for seamless migration into legacy UUID database columns.

✓

Crockford Base32 Compliance

Case-insensitive, excludes ambiguous characters (I, L, O, U) to avoid human transcription errors.

✓

Developer Code Snippets

Implementation snippets ready to copy for Node.js, Python, Go, and Rust.

Practical Engineering Scenarios

✓
Database B-Tree Index Optimization
Replace random UUIDv4 primary keys in Postgres and MySQL to eliminate cache misses and page splits.
✓
Distributed Event Sourcing & Messaging
Tag Kafka and RabbitMQ event stream payloads with naturally time-sortable IDs.
✓
Public URL & API Resource Identifiers
Expose clean, 26-character, URL-safe resource IDs (e.g. stripe-like or linear-like IDs) without leaking consecutive database auto-increments.
✓
Log & Telemetry Correlators
Embed traceable trace IDs across distributed microservices.

Frequently Asked Questions

What is the difference between a ULID and a UUIDv4?

UUIDv4 is 128 bits of pure pseudorandomness, which causes severe B-Tree index fragmentation when used as a database primary key. A ULID is also 128 bits, but its first 48 bits encode a Unix millisecond timestamp, making it lexicographically sortable and highly index-friendly.

Can ULIDs be stored in a PostgreSQL UUID column?

Yes! Because ULIDs are 128 bits in binary, they can be seamlessly converted to standard 36-character hexadecimal UUID format and stored directly in PostgreSQL's native 16-byte `UUID` data type.

What happens if two ULIDs are generated in the exact same millisecond?

Standard ULID implementations use a monotonic factory: when generating multiple IDs within the same millisecond, the random component is incremented by 1 bit, guaranteeing absolute sequential sort order without collisions.

What is Crockford's Base32 encoding?

Crockford's Base32 is an alphanumeric encoding alphabet of 32 characters (0-9, A-Z) that intentionally excludes I, L, O, and U to prevent confusion with numbers 1 and 0, and to avoid accidental obscenities.