FOCUS 1.4 · FinOps Open Cost & Usage Specification
FOCUS 1.4 Conformance
& Power BI Field Mapping
PromptKing exports native FOCUS 1.4 conformant AI billing data. Download standard FOCUS CSV or JSON and load it directly into Power BI, Tableau, or Snowflake — no custom ETL required. A FOCUS 1.3 legacy export remains available for pipelines mid-implementation.
Ratified at FinOps X 2026 · San Diego · June 4, 2026
FOCUS 1.4 — ratified at FinOps X 2026 — extends the open billing standard to AI token economics. PromptKing exports conformant FOCUS 1.4 datasets across every connected AI vendor: token quantities, service categorization, and effective cost, normalized into the industry's common language.
With the Linux Foundation's Tokenomics Foundation now co-funding the FOCUS token-economics expansion, token-level cost data is becoming the industry's shared dialect. PromptKing speaks it natively.
FOCUS 1.4 Conformant — token economics columns mapped or conformantly null
Spec: focus.finops.org/focus-specification/ · Conformance file: /focus-conformance.json · FOCUS 1.3 legacy export: ?spec=1.3
Conformance Declaration
Power BI — 4-Step Quick Start
Download the FOCUS CSV
Click "Download 1.4 CSV" above or navigate to /api/v1/focus-export?format=csv&period=30d&spec=1.4. Save the file locally.
Load into Power BI
Open Power BI Desktop → Get Data → Text/CSV → select the file. Use "Detect data types" — Power BI will auto-type dates, decimals, and booleans from the column names.
Set column types
BillingPeriodStart / BillingPeriodEnd → Date. BilledCost / EffectiveCost / ListCost → Decimal. x_IsAgentic → True/False. x_UtilisationPct → Decimal (0–1).
Create the CFO summary page
Add: Card visual (Total AI Spend = SUM[BilledCost]), Donut chart (BilledCost by ServiceProviderName), Table (SubAccountName, x_PlanType, BilledCost, x_UtilisationPct), Filter (ChargeCategory = "Purchase") for seat analysis.
Power BI DAX Measures — Copy & Paste
Create these measures in Power BI (Modeling → New Measure). Assumes your table is named FOCUS.
Core FOCUS 1.4 Columns — Cost and Usage Dataset
FOCUS 1.4's supplemental datasets (Invoice Detail, Billing Period, expanded Contract Commitment) are optional and not yet emitted — the Cost and Usage dataset below is the conformance surface for AI subscription billing. Columns with no data source in our schema are emitted null, which is conformant per FOCUS rules.
Custom AI Columns (x_ prefix — FOCUS extensibility)
Per FOCUS extensibility guidelines, custom columns use the x_ prefix. FOCUS 1.4 keeps one row per charge for AI token economics — input, output, and cached token splits ride as x_ extension columns. These expose AI-specific dimensions not in the core spec — ghost seat detection, agentic usage flagging, and model-level cost attribution.
Tableau & Snowflake
Tableau Desktop
- →Connect → Text File → select FOCUS CSV
- →Date columns (BillingPeriodStart/End) auto-detect as Date
- →Create a Vendor Spend dashboard: Rows = ServiceProviderName, Columns = MEASURE(BilledCost)
- →Add filter: ChargeCategory = "Purchase" for seat-level view
- →x_UtilisationPct < 0.2 filter surfaces ghost seats instantly
Snowflake / Data Warehouse
- →Load FOCUS JSON via Snowflake COPY INTO from stage
- →Schema: all numeric columns as FLOAT, dates as DATE, booleans as BOOLEAN
- →x_ columns map to AI-specific dimension tables
- →JOIN on SubAccountId for seat-level enrichment from HR/identity systems
- →Recommended: partition by BillingPeriodStart for query performance
Ready to connect your AI vendors?
Connect Anthropic, GitHub Copilot, Google Workspace, M365, Bedrock, or Watsonx — and export live FOCUS data to your BI stack in minutes.