Create your first database
Provision a Silos database in seconds from the Console, the CLI, or the REST API.
A Silos database is "just data" — a set of metadata plus snapshot and WAL storage — until something queries it. That's what makes provisioning fast: creating a database doesn't spin up a server, it allocates storage and metadata. You can create one three ways.
Create a database by name:
silos db create my-appList your databases to confirm it's there:
silos db listWhen you're done with a database, you can delete it just as quickly — which is exactly the lifecycle that suits ephemeral, per-tenant, or agent-driven workloads.
From the dashboard at silos.sh:
Open the dashboard
Sign in and go to your project dashboard.
Create a database
Click New database, give it a name, and create it. The database appears in your list immediately.
Grab your connection details
Open the database to find its connection string and management options — branches, metrics, and the SQL editor.
Silos exposes a Neon-shaped REST API. Create a database under a project with a request shaped like:
curl -X POST "https://api.silos.sh/v1/projects/{project_id}/databases" \
-H "Authorization: Bearer $SILOS_API_KEY" \
-H "Content-Type: application/json" \
-d '{ "database": { "name": "my-app" } }'This is the path AI agents and CI pipelines use to provision databases programmatically. See the API reference for the full surface.
What you get
Each database comes with a primary branch named main and a connection endpoint of
the form:
postgres://user:password@db-xxxx.silos.sh/main?sslmode=requireThe db-xxxx host is unique to your database. Connections require TLS — sslmode=require
is not optional.
A new database scales to zero when idle and cold-starts on its first query. The very first connection after a period of inactivity pays the materialization cost; see Scale to zero.