Anyscale · FinOps Profile

Anyscale Finops

FinOps view of Anyscale spend. Anyscale meters compute by instance-hour per node type (CPU and NVIDIA T4 / L4 / A10G / A100 / H100 / H200). Hosted customers are invoiced directly by Anyscale; BYOC customers are charged the underlying cloud spend by AWS / GCP plus an Anyscale platform fee per instance-hour, often via a cloud marketplace.

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

It defines 8 billable meters, billed in USD, on a monthly cycle, and pricing category usage-based.

The profile maps 8 FOCUS columns for cost-allocation reporting.

Tagged areas include AI, Distributed Computing, Ray, ML Platform, and Inference.

Category: AI and Machine Learning Pricing: Usage-Based Billing: Monthly FOCUS v1.3
AIDistributed ComputingRayML PlatformInferenceGPUFinOpsCost ManagementFOCUS

Framework Alignment

Framework
FinOps Foundation Framework
Data Spec
FOCUS v1.3

Charge Categories

UsagePurchaseAdjustment

FOCUS Columns

BillingCurrency
USD
ChargeCategory
Usage
InvoiceIssuerName
Anyscale
PricingCategory
Usage-Based
ProviderName
Anyscale
PublisherName
Anyscale
ServiceCategory
AI and Machine Learning
ServiceName
Anyscale Platform

Meters

cpu_instance_hours
Unit: hours
CPU-only node hours billed per instance-hour ($0.0135/hr hosted).
gpu_instance_hours_t4
Unit: hours
NVIDIA T4 GPU instance hours ($0.5682/hr hosted).
gpu_instance_hours_l4
Unit: hours
NVIDIA L4 GPU instance hours ($0.9542/hr hosted).
gpu_instance_hours_a10g
Unit: hours
NVIDIA A10G GPU instance hours ($1.3635/hr hosted).
gpu_instance_hours_a100
Unit: hours
NVIDIA A100 GPU instance hours ($4.9591/hr hosted).
gpu_instance_hours_h100
Unit: hours
NVIDIA H100 GPU instance hours ($9.2880/hr hosted).
gpu_instance_hours_h200
Unit: hours
NVIDIA H200 GPU instance hours ($10.6812/hr hosted).
byoc_platform_hours
Unit: hours
BYOC platform-fee instance-hours invoiced via marketplace (rate contract-dependent).

Sources