Nvidia · FinOps Profile

Nvidia Finops

FOCUS-aligned FinOps for NVIDIA: hybrid commercial model spanning per-GPU/year AI Enterprise subscriptions (sold direct or via channel), hourly cloud marketplace SKUs (AWS/Azure/GCP), DGX Cloud hosted GPU rentals, and free build.nvidia.com trial credits. Cost allocation is GPU-centric: licenses, cloud-instance hours, and inference token throughput.

Nvidia Finops is the FinOps profile for Nvidia on the APIs.io network, aligned with the FinOps Foundation Framework.

It defines 5 billable meters, billed in USD, on a annual (subscription) or hourly (cloud) cycle, and pricing category subscription (per-gpu) + pay-as-you-go (cloud marketplace) + free trial.

The profile maps 9 FOCUS columns for cost-allocation reporting.

Tagged areas include GPU, AI, Machine Learning, Computing, and Graphics.

Category: AI Infrastructure Pricing: Subscription (Per-GPU) + Pay-As-You-Go (Cloud Marketplace) + Free Trial Billing: Annual (subscription) or Hourly (cloud) FOCUS v1.3
GPUAIMachine LearningComputingGraphicsFinOpsCost ManagementFOCUS

Framework Alignment

Framework
FinOps Foundation Framework
Data Spec
FOCUS v1.3

Charge Categories

PurchaseUsageTaxCreditAdjustment

FOCUS Columns

BillingCurrency
USD
ChargeCategory
Purchase
InvoiceIssuerName
NVIDIA Corporation (or AWS/Azure/GCP for marketplace)
PricingCategory
Subscription
PricingUnit
gpu
ProviderName
NVIDIA
PublisherName
NVIDIA Corporation
ServiceCategory
AI Infrastructure
ServiceName
NVIDIA AI / Developer Platform

Meters

ai_enterprise_gpu_subscription
Unit: gpu-year
Per-GPU per-year AI Enterprise software subscription
ai_enterprise_marketplace_hours
Unit: gpu-hour
AI Enterprise hourly metered consumption on cloud marketplaces
dgx_cloud_instance_hours
Unit: instance-hour
DGX Cloud / DGX Cloud Lepton hosted GPU instance hours
build_api_credits
Unit: credit
build.nvidia.com NIM endpoint credit consumption
inference_tokens
Unit: token
Inference tokens generated (input/output) on hosted or self-hosted NIM (operational telemetry)

Sources