NanoID & ULID Generator
Generate cryptographically secure, URL-safe NanoIDs and time-sortable ULIDs in batches. Calculate collision probabilities and inspect timestamp metadata client-side.
~4.1 Trillion Years
Until 1% Probability of Collision
ULID Inspector & Timestamp Decoder
Paste any ULID below to extract its exact millisecond generation timestamp and randomness entropy.
NanoID vs ULID vs UUID v4 vs UUID v7
| Identifier | Standard Length | Character Set | URL Safe? | Lexicographically Sortable? | Collision Resistance |
|---|---|---|---|---|---|
| NanoID | 21 chars (configurable 6–64) | Customizable (URL-safe A-Za-z0-9_-) | Yes | No (Random) | High (~Trillions of years) |
| ULID | 26 chars | Crockford's Base32 (no I, L, O, U) | Yes | Yes (1ms precision) | 128-bit (80 bits random entropy) |
| UUID v4 | 36 chars (with hyphens) | Hexadecimal (0-9, a-f) | Hyphenated | No (Random) | 122-bit pseudo-random |
| UUID v7 | 36 chars (with hyphens) | Hexadecimal (0-9, a-f) | Hyphenated | Yes (Unix Epoch) | 128-bit time + random |
Frequently Asked Questions
Why choose NanoID over standard UUID v4?
NanoID is shorter (21 characters vs 36 characters), contains no awkward hyphens, uses a larger 64-character URL-safe alphabet, and is completely cryptographically secure. It takes up less database indexing storage and renders cleanly in URLs without percent-encoding.
What makes ULID ideal for databases like Postgres and MongoDB?
Standard random UUIDs cause B-Tree index fragmentation when inserted in high volumes because consecutive keys are scattered randomly across disk pages. ULIDs begin with a 48-bit millisecond timestamp, making them naturally monotonically increasing. This preserves sequential disk writes and dramatically improves database insert throughput.
Is ID generation secure and private?
Yes. All IDs are created using the browser's hardware-backed Web Cryptography API (
window.crypto.getRandomValues()). None of your generated IDs are transmitted over the internet or logged by servers.