CUDO Compute · Pricing Plans

Cudo Compute Plans Pricing

CUDO Compute uses a usage-based, pay-as-you-go model billed per GPU-hour, per vCPU-hour, per GB of memory-hour, and per GB of storage, with rates that vary by GPU model and data center. On-demand VMs are billed by the second/hour; Reserved Capacity and committed contracts unlock lower rates across multiple term lengths. Bare metal and multi-node GPU clusters are typically priced by quote/reservation. The API exposes live machine-type prices per data center.

Cudo Compute Plans Pricing is the machine-readable pricing-plan profile for CUDO Compute on the APIs.io network, conforming to the API Commons Plans specification.

It defines 3 plans, covering usage, subscription, and enterprise tiers, with named plans including On-Demand, Reserved Capacity, Bare Metal & Clusters.

Tagged areas include GPU, Cloud Compute, Infrastructure, Virtual Machines, and Marketplace.

3 Plans API Commons Plans
View Source
GPUCloud ComputeInfrastructureVirtual MachinesMarketplacePlans

Plans

On-Demand usage

Pay-as-you-go GPU and CPU virtual machines billed per resource-hour with no commitment; rates vary by GPU model and data center.

NVIDIA H100 (per GPU) (gpu_hours · hour) ~$1.80-$2.45 per GPU/hr (representative; varies by data center) USD
NVIDIA H200 (per GPU) (gpu_hours · hour) quote on request / per data center USD
NVIDIA A100 80GB (per GPU) (gpu_hours · hour) ~$1.50-$1.60 per GPU/hr (representative; varies by data center) USD
NVIDIA L40S (per GPU) (gpu_hours · hour) per data center (see live catalog) USD
NVIDIA V100 (per GPU) (gpu_hours · hour) per data center (see live catalog) USD
NVIDIA A800 (per GPU) (gpu_hours · hour) per data center (see live catalog) USD
vCPU (vcpu_hours · hour) per vCPU/hr (varies by CPU model and data center) USD
Memory (gib_hours · hour) per GiB/hr (varies by data center) USD
Block Storage (gib_month · month) per GiB/month (varies by storage class and data center) USD
Reserved Capacity subscription

Commitment-based capacity at discounted rates across multiple term lengths, with capacity guarantees for GPU VMs, dedicated servers, and clusters.

Committed GPU Capacity (gpu_hours · month) discounted vs. on-demand (term-based; contact sales) USD
Bare Metal & Clusters enterprise

Dedicated bare metal servers and multi-node GPU clusters (H200, H100, A100 80GB, L40S, A800, and Blackwell-generation GPUs) priced by quote and reservation with InfiniBand/high-bandwidth networking options.

Cluster Reservation (contract · month) quote on request USD

Sources