---
title: Regions
description: Where a resource's data lives, how placement is declared, and which locations answer today.
type: explanation
product: platform
summary: A region is a string you set on a resource; one location answers today.
updated: 2026-09-28
order: 5
---

A region says where a resource's data lives. You set it on the resource itself,
so two resources in the same project can live in different places.

## Where placement is declared

Two surfaces carry placement:

- The resource's own spec, for anything you create one at a time: a database's
  `region`, a service's `regions`, a bucket's replicated regions, a key's
  regions.
- The app manifest's placement intent, for everything an app deploys at once:
  `locationIds`, `residency`, `jurisdiction`, `latencyClass` and
  `failureDomain`.

A location id is a string that starts with `loc_`. A service that names no
region of its own is served in the home location of its project.

Inside a region, resources are served by a cell: one isolated serving unit. A cell
has a four-character code that never changes, for example `fs01`. It appears
in resource ids (`db_fs01…`) and in the `cell`, `cells` and `home_cell`
fields. Placement chooses the cell; a release may name specific cells.

## What is live today

One location answers today: Falkenstein, Germany, whose region slug is `fsn1`.
A resource that names no particular region is served there. More locations are
added as the platform grows, and the placement fields above are the ones that
will name them.

<Callout tone="note" title="Region is data, not a switch">
	The placement fields are carried as plain strings. Treat the value you set as
	data about your resource, keep it in your configuration, and read the
	location's own name back from the resource rather than assuming it.
</Callout>

<RelatedDocs
	links={[
		{ href: '/docs/platform/accounts', label: 'Accounts and projects' },
		{ href: '/docs/platform/environments', label: 'Environments' },
	]}
/>