UUID Generator - Click to Decode Every Segment
Generate UUIDs in every version - v1, v3, v4, v5, v7 - and click any part of the UUID to see exactly what it encodes. Bulk-generate up to 10,000, decode timestamps, validate existing UUIDs, and export as JSON, SQL, or CSV.
What makes it different
Other generators give you
random text. This explains it.
Click into any UUID to see what it encodes. Compare versions. Investigate existing UUIDs. Generate thousands with real export formats.
Click-to-decode segments
Every part of the UUID is interactive - click any segment to see exactly what it encodes: timestamp, version, variant, or node ID.
the UUID itself is the UIGenerate & compare diff
See the previous UUID next to the new one with differences highlighted - instantly shows why v1/v7 look sequential while v4 looks fully random.
visual proof of randomnessInvestigate mode
Paste any UUID and watch it get analysed like a terminal log - version detected, timestamp extracted, namespace identified.
terminal-style revealBulk generation, real exports
Generate up to 10,000 UUIDs with a live virtualized table, then export as JSON, SQL INSERT, CSV, or a Python list - ready for real workflows.
seed data, instantlyNamespace determinism proof
Generate the same v3/v5 UUID twice and watch it match exactly - then change one character and watch it completely change.
prove it to yourselfCollision probability calculator
Set your own generation rate and timeframe to see exactly how astronomically unlikely a UUID collision really is.
satisfies real curiosityQuick guide
Generate or decode a UUID in seconds
Pick a version
v4 for general use, v1/v7 for timestamp-based, v3/v5 for deterministic namespace-based UUIDs.
Click to explore
Click any part of the generated UUID to see what it encodes. Switch to Bulk for thousands at once.
Copy or export
Copy a single UUID in any format, or export bulk results as JSON, SQL, CSV, or a Python list.
UUIDs - what's actually inside one
A UUID looks like random noise, but it has real structure. It is 128 bits, written as 32 hexadecimal digits grouped into five sections: 8-4-4-4-12 characters, separated by hyphens. The third group always starts with a digit that identifies the UUID's version, and the fourth group's first character encodes the variant (almost always RFC 4122).
Choosing the right version
| Version | How it's built | Best for |
|---|---|---|
| v1 | Timestamp + MAC address | Legacy systems, when sortability matters and v7 isn't available |
| v3 | MD5 hash of namespace + name | Deterministic IDs, when MD5 collision resistance isn't a concern |
| v4 | Fully random (122 random bits) | General purpose - the most common choice today |
| v5 | SHA1 hash of namespace + name | Deterministic IDs - preferred over v3 |
| v7 | Timestamp (48-bit) + random bits | Database primary keys - sortable and unique |
Namespace UUIDs - same input, same output, every time
Unlike v1/v4/v7, versions 3 and 5 are deterministic: hashing the same namespace UUID and name string will always produce the identical UUID. This is useful when you need a stable identifier derived from existing data - for example, generating a consistent UUID for a domain name without having to store a separate random ID and look it up later. RFC 4122 defines four standard namespaces: DNS, URL, OID, and X.500.
Are UUID collisions actually possible?
Technically yes, but the probability is so small it's treated as zero in practice. A UUID v4 has 122 random bits (6 bits are fixed for version and variant). Using the birthday paradox approximation, you would need to generate roughly 2.71 quintillion UUIDs to have a 50% chance of a single collision. Try the Collision Odds tab above with your own numbers to see this for yourself.