RemoteGPU Pricing
See current pricing for Comfy, Inference API, Kubernetes, Virtual Machine, and Storage.
Hosted workspace
Comfy
Build and run Comfy workflows in the browser.
Hosted Comfy H100 x1-----
Hosted Comfy H100 x2-----
Hosted Comfy H100 x4-----
Hosted Comfy H100 x8-----
Models & endpoints
Inference API
Connect your app to model endpoints.
Qwen/Qwen3.6-27B-FP8--
Compute Infrastructure
Kubernetes
Deploy containers, jobs, and services with Kubernetes.
Kubernetes H100 x1-----
Kubernetes H100 x2-----
Kubernetes H100 x4-----
Kubernetes H100 x8-----
Compute Infrastructure
Virtual Machine
Launch GPU and CPU virtual machines with direct guest access.
Virtual Machine H100 x1-----
Virtual Machine H100 x2-----
Virtual Machine H100 x4-----
Virtual Machine H100 x8-----
Persistent volumes
Storage
Keep models, files, and outputs across sessions and pods.
Standard Block Storage8 GB--
Standard Block Storage16 GB--
Standard Block Storage32 GB--
Standard Block Storage64 GB--
Standard Block Storage128 GB--
Standard Block Storage256 GB--
Standard Block Storage512 GB--
Standard Block Storage1 TB--
Standard Block Storage2 TB--
Billing notes
- Published rates are billed through Stripe per reserved hour.
- Storage is shown per reserved capacity hour.
- Hourly prices are rounded for display; invoices use exact Stripe metered rates.
- Image inference is billed per compute unit; final usage varies by model, runtime, output size, input images, and request settings.
- Monthly estimates use 730 hours.