Price before deployment
The selected GPU and capacity type determine the hourly compute rate shown before creation.
GPU Cloud
Deploy containerized GPU instances from one workspace.
Choose an OCI image, Paris or Frankfurt request region, persistent storage, capacity mode and resource settings, then create and manage the configuration from the dashboard.
Compare GPU hardware and catalog ratesTransparent baseline
$0.16/hr
NVIDIA RTX A5000 on-demand from
19
Priced GPU models
OCI
Container images
2
EU request regions
1
Workspace control plane
EU regions, encrypted storage and network isolation options
Why GPURento
Less infrastructure work. More model progress.
Bring a container image
Use your OCI image, persistent volume size and resource name in one repeatable configuration.
Choose an EU region
Select Paris or Frankfurt as the requested execution region from the same workspace.
From code to capacity
One clean path to production.
1
Pick capacity
Filter by GPU, VRAM, region, uptime tier and total cost.
2
Attach your environment
Select a template or point to a private container registry.
3
Create from the dashboard
Prepaid resources need a positive wallet balance. The $50 minimum applies to deposits, not to your remaining balance.
deploy.py
await fetch("/api/instances", {
method: "POST",
headers: { "content-type": "application/json" },
body: JSON.stringify({
name: "training-h100",
gpuId: "h100-sxm",
region: "eu-west-par",
image: "pytorch/pytorch:2.6-cuda12.6",
capacityType: "on-demand",
volumeGb: 200
})
})✓ 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.
On-demand and spot modes
19 priced GPU configurations
Paris and Frankfurt regions
Persistent volume size per resource
Container image configuration
Workspace ownership checks
Wallet funding gate
Resource and usage records
What is a cloud GPU instance?+
It is a dedicated, containerized compute environment with direct access to one or more GPUs. You control the image, libraries, storage and exposed services.
Can I use my own Docker image?+
Yes. The resource stores an OCI-compatible image reference along with its storage and region settings.
What remains billable when an instance stops?+
Compute stops accruing, while persistent volumes and reserved IPs can continue to incur storage or allocation charges. The console shows both running and stopped-state estimates before deployment.
Which GPU should I choose?+
Use VRAM first, then expected throughput and budget. The GPU directory compares memory, architecture, pricing and recommended workloads.
Ready when you are