Understanding UUIDv7: Time-Sortable UUIDs

The modern UUID format with embedded timestamps for better database performance

UUIDv7 is a new universally unique identifier format defined in RFC 9562 (May 2024). Unlike random UUIDs (v4), UUIDv7 embeds a Unix timestamp in milliseconds, making the identifiers time-sortable while maintaining global uniqueness.

UUIDv7 Structure

01936b2e-1e85-7000-8000-000000000000

48-bit timestamp (ms)

Version (7)

Variant (8, 9, a, b)

Random bits

Did you know?ISO 8601 is the international standard for date formatsLearn more

How UUIDv7 Works

A UUIDv7 consists of 128 bits organized as follows:

Bits 0-47Unix timestamp in milliseconds since epoch

Bits 48-51Version indicator (always 7)

Bits 52-63Random or counter data for sub-millisecond ordering

Bits 64-65Variant indicator (RFC 4122 format)

Bits 66-127Random data for uniqueness

Why Use UUIDv7?

Time-Sortable

Natural chronological ordering without additional timestamp columns

Better Index Performance

Sequential inserts reduce B-tree fragmentation in databases

Globally Unique

No coordination needed between distributed systems

Embedded Timestamp

Extract creation time directly from the ID itself

UUIDv7 vs. Other UUID Versions

Version	Sortable	Timestamp	Best For
UUIDv1	Partially	Yes (100ns)	Legacy systems
UUIDv4	No	No	Random IDs
UUIDv7	Yes ✓	Yes (ms) ✓	Database PKs
ULID	Yes	Yes (ms)	UUID alternative

Did you know?The Y2038 problem affects 32-bit systemsLearn more

Database Performance Benefits

When using random UUIDs (v4) as primary keys, each new record is inserted at a random position in the B-tree index, causing:

✗Index page splits and fragmentation
✗Poor cache utilization
✗Slower write performance over time

UUIDv7 solves this because new records are always appended near the end of the index (they're time-ordered), similar to auto-incrementing integers but with the benefits of UUIDs.

Extracting Timestamps from UUIDv7

1Remove the hyphens from the UUID

2Take the first 12 hexadecimal characters

3Convert from hexadecimal to decimal

4Result is a Unix timestamp in milliseconds

JavaScript

function extractTimestamp(uuidv7) {
  const hex = uuidv7.replace(/-/g, '').slice(0, 12);
  return parseInt(hex, 16);
}

const uuid = '01936b2e-1e85-7000-8000-000000000000';
const timestamp = extractTimestamp(uuid);
console.log(new Date(timestamp));
// Output: 2024-12-01T10:30:45.000Z

Python

from datetime import datetime

def extract_timestamp(uuidv7: str) -> int:
    hex_str = uuidv7.replace('-', '')[:12]
    return int(hex_str, 16)

uuid = '01936b2e-1e85-7000-8000-000000000000'
timestamp = extract_timestamp(uuid)
print(datetime.fromtimestamp(timestamp / 1000))
# Output: 2024-12-01 10:30:45

Generating UUIDv7 in Code

Node.js 20+

import { randomUUID } from 'crypto';

// Node.js 20+ supports UUIDv7 natively
const uuid = crypto.randomUUID({ type: 'v7' });

Python

# pip install uuid7
from uuid_extensions import uuid7

uuid = uuid7()
print(str(uuid))

PostgreSQL 17+

-- PostgreSQL 17+ supports UUIDv7
SELECT gen_random_uuid_v7();

When to Use UUIDv7

✓ Ideal For

Database primary keys — especially high-write scenarios
Distributed systems — no coordination needed
Event sourcing — natural event ordering
Audit logs — embedded timestamp in every ID
APIs — client-generated, chronological IDs

⚠ Consider Alternatives

Privacy concerns — IDs reveal creation time (use UUIDv4)
64-bit constraint — use Snowflake IDs instead
Human-readable — use NanoID or custom schemes

Try Our UUIDv7 Converter

Extract timestamps from UUIDv7 identifiers and convert to human-readable dates.

UUIDv7 → Human Date Unix → Human Date

What is Unix Time?

A complete guide to epoch timestamps and why developers use them