FinOps X 2026 · June 8-11 · San Diego
Register Now
FOCUS for AI
Technology Categories / FOCUS for AI

AI usage resembles the resource consumption already found in public cloud, SaaS, and similar billing data, but it adds abstracted meters not tied to hardware: tokens, API calls, and outcomes. These rely on data generators to report what was used, each with its own reconciliation mechanisms and often bespoke ways of capturing usage internally.

Giving AI Billing a Common Shape

FOCUS has the building blocks to help FinOps Practitioners map AI charges into the same columns as everything else so they’re queryable side by side.

AI Usage FOCUS Columns
1,000,000 input tokens ConsumedQuantity + ConsumedUnit
The model & price point SkuId, SkuPriceId, PublisherName
It’s an AI service ServiceCategory: AI and ML
Paid in credits / tokens PricingCurrency → BilledCost
Reserved throughput CommitmentDiscount columns

Several public cloud data generators already report AI service usage in tokens and by SKU, as do data clouds such as Snowflake. AI-specific providers are adopting the spec too, including Nebius, which delivers FOCUS-formatted usage data for AI services.