rp stock dc
List datacentres with GPU stock and S3-API support.
rp stock dc [--json] [--s3] [--global-network] [--volume-type <t,…>]
[--compliance <c,…>] [--region <r,…>]
Options
--json print the raw v2 datacentre records (post-filter)
--s3 keep only S3-API-enabled datacentres
--global-network keep only datacentres on Runpod's global network
--volume-type <t,…> keep only datacentres supporting at least one listed
network-volume tier (e.g. STANDARD,HIGH_PERFORMANCE)
--compliance <c,…> keep only datacentres carrying at least one listed
certification (e.g. SOC_2_TYPE_2)
--region <r,…> keep only datacentres in at least one listed region
(e.g. EU,NA,AS) — the REGION column, case-insensitive.
Accepts abbreviations (eg: NA, EU) or the full names
(eg: NORTH_AMERICA, ASIA)
Notes
Filters combine as a logical AND; within a comma list they are OR. All flags
apply to BOTH the table and --json, so the two views always show the same
datacentres (--json omits the S3_API column, which is a CLI-side join).
The DATACENTER column is the id rp pod create --dc and
rp volume create --dc take. (The v2 name field is the same value in
practice and is not shown as a separate column.)
GPUS counts how many GPU types have any stock there, not how many cards
are free.
GLOBAL_NETWORK marks datacentres that support global networking — Runpod's
private, cross-datacenter pod-to-pod network (pods reach each other over
*.runpod.internal without opening ports to the public internet). It is
unrelated to the S3_API column.
COMPLIANCE lists the certifications the datacentre carries (e.g.
SOC_2_TYPE_2), comma-separated; blank where none are advertised.
NETWORK_VOLUME_TYPES lists the network-volume tiers the datacentre supports
(e.g. STANDARD, HIGH_PERFORMANCE), comma-separated; blank where none are
advertised. The tier constrains the kind of volume rp volume create can
place there.
S3_API marks the datacentres whose network volumes expose the
S3-compatible API — the ones rp volume sync can reach.
That column is NO-V2-EQUIVALENT: v2 carries no S3 field anywhere, so it is
joined in from the GraphQL dataCenters query — the same resolver
rp volume create guards on — with an offline snapshot behind it. The
column therefore stays live where GraphQL is reachable and still renders
when it is not.
--json prints the v2 datacentre records alone: S3_API is a CLI-side join
and is absent from that payload.
Examples
# Datacentres that are BOTH S3-API enabled AND on the global network
$ rp stock dc --s3 --global-network
# Datacentres that support the HIGH_PERFORMANCE volume tier
$ rp stock dc --volume-type HIGH_PERFORMANCE
API: GET /v2/catalog/datacenters (include=GPU_AVAILABILITY,CPU_AVAILABILITY)