Network Storage

Your data outlives every GPU.

A network volume is persistent storage that exists independently from compute. Keep datasets, checkpoints and model artifacts, then attach the volume to a cluster in the same region.

Transparent baseline

$0.07/GB/mo

standard NVMe up to 1 TB

Self-service

4 TB

Self-service maximum

$0.05

Per GB over 1 TB

$0.14

High performance / GB

2

EU regions

EU regions, encrypted storage and network isolation options

Why GPURento

Less infrastructure work. More model progress.

Independent from compute

Deleting or replacing a GPU resource does not remove its standalone network volume.

Region-scoped

Create storage in Paris or Frankfurt and attach it only to compute in the same region.

Grow when needed

Increase a volume up to 4 TB from the dashboard while preserving its mount path and history.

From code to capacity

One clean path to production.

1

Choose class and size

Select Standard or High Performance NVMe, region, mount path and 10–4,000 GB.

2

Create after funding

Prepaid resources need a positive wallet balance. The $50 minimum applies to deposits, not to your remaining balance.

3

Attach to a cluster

Pick the volume during cluster creation; it remains reusable after that cluster is removed.

deploy.py
await fetch("/api/storage", {
  method: "POST",
  headers: { "content-type": "application/json" },
  body: JSON.stringify({
    name: "training-data",
    region: "eu-west-par",
    storageClass: "standard",
    sizeGb: 500,
    mountPath: "/workspace"
  })
})
✓ Configuration saved to the workspace · provisioning state visible

Included

A serious platform, not a thin wrapper.

Every workload gets the same API, identity, billing and observability layer.

Persistent network volumes
Standard NVMe storage
High performance NVMe storage
Paris and Frankfurt regions
10 GB to 4 TB self-service
Grow without recreating
Cluster attachment tracking
Workspace audit history

FAQ

Straight answers.

Read the docs
What is network storage?+

It is a persistent volume whose lifecycle is separate from GPU compute. It is designed for reusable datasets, checkpoints, weights and artifacts.

Can users create storage without contacting sales?+

Configure resources directly in the dashboard. Prepaid usage uses wallet credit; monthly rentals have a separate checkout.

Can I reduce a volume?+

No. A volume can grow but cannot shrink, which avoids truncating existing data.

Can a Frankfurt cluster attach a Paris volume?+

No. The volume and cluster must share the same region.

Ready when you are

Ship the workload, not the wait.

Create account