GPU Marketplace
NVIDIA P100 PCIeOn-Demand
NVIDIA GeForce RTX 3070On-Demand
NVIDIA GeForce RTX 3080On-Demand
NVIDIA GeForce RTX 3090On-Demand
NVIDIA GeForce RTX 4090On-Demand
NVIDIA RTX A6000On-Demand
Estimates only — rates are collected automatically from public provider pages and may be out of date. Prices vary by region, commitment term, and availability, and typically exclude storage, egress, and tax. Confirm current pricing with the provider before purchasing. Last collected .
Company Profile
Infrastructure
GPU FleetGPU types include examples such as H100_80GB and A100_80GB
Compute & Deployment
On-DemandThe price is defined in `prices.onDemand`.
GPU Hardware
Multi-GPU NodesSupported (API exposes gpu_count and references multi-GPU setups)
NVLinkInterconnect types may include NVLink
InfiniBandInterconnect types may include InfiniBand
PCIe vs SXMSupports PCIe and SXM variants (SXM2, SXM3, SXM4, SXM5)
Pricing Model
Performance & Scaling
Multi-Node TrainingisMultinode: Indicates whether this disk can be attached to multiple instances simultaneously. This is useful for shared data scenarios where multiple GPU instances need access to the same dataset.
InfiniBandFor multi-GPU setups, these fields specify the interconnect speed (in Gbps) and technology type (e.g., `InfiniBand`, `NVLink`) used for GPU-to-GPU communication.
Developer Experience
OnboardingRequires an API Key with Availability -> Read permission; API requests require an Authorization: Bearer token with your API key; CLI requires setting API keys and SSH key path as outlined in the CLI introduction.
CLI ToolingProvides a CLI tool 'prime' with commands such as 'prime availability list' and 'prime availability disks' and supports filters like --gpu-type, --gpu-count, --regions, --socket, --disks, --data-center-id, --cloud-id.
TemplatesProvides prebuilt images such as ubuntu_22_cuda_12, cuda_12_1_pytorch_2_2, cuda_11_8_pytorch_2_1, stable_diffusion, flux, axolotl, bittensor, vllm_llama_8b, vllm_llama_70b, vllm_llama_405b
DocumentationContains API reference for availability and disks, example curl requests and JSON responses, a response schema reference link, an interactive Availability API for request samples, and a CLI reference with commands and flags.
Security & Compliance
Responses include a 'secure_cloud' security designationexample provider names shown include 'runpod' and 'hyperstack'.
Data Center Locations
Coverage
CountriesUnited States, Canada, US, NO, NORWAY
Filter by region(s) (e.g., `united_states`, `canada`, `europe_west`)
Compliance Regions
Datacenter Locations
Key Strengths
Provides a dedicated Availability API (GPU and disk endpoints) with rich filtering (region, gpu_count, gpu_type, socket, security, data_center_id, cloud_id, disks), ability to find GPUs that can attach to users' existing disks, returns provider/cloudId required for provisioning, offers interactive request samples and a CLI.
Known Limitations
Some response fields may be null or unspecified in examples: provisioningTime may be null
isSpot may be null
prepaidTime may be null
disk cloudId may be null.
Additional Information
Last updated August 2026. Information subject to change.