[Upcoming] UUIDv7 Adoption

In the next couple of months, we intend to add support for UUIDv7 in Primary Key (PK) fields. During the new data model creation process, you will be able to choose the type of PK between the old Int and the new UUIDv7.

UUIDv7 support will bring significant improvements to many complex projects. While UUIDv7 has some overhead compared to Int, it will definitely a nice addition:

  • No issues with sequences, like with Int, when importing data into tables.
  • You can generate UUIDs in your application without relying on DB-generated IDs if desired.
  • Ideal for building distributed systems: every app can generate its own IDs without collisions.
  • UUID significantly improves the ability to partially restore database records from backups in the event of a disaster.

Additionally, we have already begun adopting UUID in Studio with a single goal: enabling record restoration in our databases and ID generation within applications before DB inserts. You may already notice some IDs in the format 01939541-a6b9-7929-9691-b95711bb9d1b. The first few characters clearly indicate UUIDv7 (019395).

PostgreSQL 17.3 will bring built-in support for generating UUIDv7 IDs. Until then, we use our implementation of the same algorithm.

You can read more about UUIDv7: