What is UUID Generator?

Creates universally unique identifiers in v4 (random), v7 (time-ordered), v1 (timestamp plus node ID), v8 (custom layout), and v5 (namespace-based and reproducible) formats. Generate single or bulk UUIDs, validate any UUID someone hands you, or derive stable IDs from a namespace and name — useful for database keys, API tokens, test data, and distributed systems.

Pick v4 for fully random IDs, v7 when you want IDs that sort by creation time (which keeps database indexes happy and makes log timelines easy to scan), v1 for a classic timestamp-plus-node ID, or v8 for the RFC 9562 custom layout. Choose v5 for deterministic IDs derived from a namespace and a name string. The validator panel decodes any UUID someone gives you and shows its version, variant, and embedded timestamp where present. Generate up to 1000 at once, toggle uppercase, copy individual lines or the whole batch, and export as plain text, spreadsheet CSV, JSON array, or ready-to-run SQL INSERT statements.

How to use

1. Step 1 — Pick a mode: v4 for fully random, v7 for time-sortable, v1 for a timestamp-plus-node ID, v8 for the custom/experimental layout, v5 for deterministic namespace-based IDs, or Validate to check an existing UUID.
2. Step 2 — For v4/v7 set how many you need (1 to 1000); for v5 choose a namespace and type the name to hash; for Validate paste the UUID and press Check.
3. Step 3 — Copy individual UUIDs or the entire batch, toggle uppercase if needed, and download as plain text, CSV, or JSON.

When to use

• Seeding a test database with primary keys before the first migration runs.
• Generating API tokens or device identifiers for distributed systems.
• Creating session IDs, trace IDs, or correlation IDs for log aggregation.

Result

You're seeding a test database and need 50 unique primary keys. Select UUID v4, set quantity to 50, generate, and copy all — each one is cryptographically random with negligible collision risk.

FAQ

What's the difference between UUID v4, v5, and v7?

v4 is 122 bits of randomness with no structure. v5 hashes a namespace UUID with a name string using SHA-1, so the same inputs always produce the same UUID — handy for deterministic IDs from emails or URLs. v7 puts a 48-bit millisecond timestamp at the start, then random bits, so a list of v7s sorts chronologically. That makes it great for database keys where insertion order matters.

How likely is a UUID collision?

Generating one billion v4 UUIDs per second for 85 years gives roughly a 50% chance of one duplicate. In normal application volumes, treat collisions as impossible without extra coordination.

Should I use a UUID as a database primary key?

Yes for distributed systems where coordinating sequential IDs is painful. Prefer v7 over v4 — random v4 keys cause index fragmentation in B-trees, whereas v7's time prefix keeps insertions clustered.

Is a UUID safe to expose in a public URL?

v4 is unguessable, so it works as a soft secret for share links or password reset tokens. v7 leaks the creation time, which sometimes matters — for purely opaque public IDs, prefer v4.

Can I shorten a UUID for friendlier URLs?

Re-encode the 128 raw bits as base32 or base62 and you get a 22 to 26-character string. Same uniqueness, less visual noise. The dashes in the standard form are just formatting and can be dropped.

Related Tools