Hugging Face · FinOps Profile

Hugging Face Finops

FOCUS-aligned FinOps for Hugging Face: per-seat subscriptions plus pay-as-you-go inference credits, per-instance-hour for Inference Endpoints and Spaces, and per-TB-month storage with volume discounts.

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

It defines 6 billable meters, billed in USD, on a monthly cycle, and pricing category subscription + pay-as-you-go.

The profile maps 7 FOCUS columns for cost-allocation reporting.

Tagged areas include FinOps, FOCUS, AI, Machine Learning, and Inference.

Category: AI Infrastructure Pricing: Subscription + Pay-As-You-Go Billing: Monthly FOCUS v1.3
FinOpsFOCUSAIMachine LearningInference

Framework Alignment

Framework
FinOps Foundation Framework
Data Spec
FOCUS v1.3

Charge Categories

UsagePurchaseCreditAdjustmentTax

FOCUS Columns

BillingCurrency
USD
ChargeCategory
Usage
InvoiceIssuerName
Hugging Face, Inc.
ProviderName
Hugging Face
PublisherName
Hugging Face, Inc.
ServiceCategory
AI Infrastructure
ServiceName
Hugging Face

Meters

subscription_seats
Unit: seat
PRO ($9/mo), Team ($20/seat/mo), Enterprise ($50+/seat/mo) seat fees.
inference_provider_credits
Unit: USD
Pay-as-you-go credits consumed against partner inference providers; pass-through provider rate, no HF markup.
inference_endpoint_hours
Unit: instance-hour
Replica-hours for dedicated Inference Endpoints, billed by hardware tier.
spaces_hardware_hours
Unit: instance-hour
Hourly cost of Spaces hardware (CPU upgrade, T4, L4, A10G, A100, etc).
hub_storage
Unit: GB-month
Tiered TB-month storage for public and private repos with volume discounts at 50/200/500 TB.
zerogpu_seconds
Unit: second
H200 GPU-seconds consumed under the ZeroGPU shared pool (free tier with quotas by plan).

Sources