Breadcrumbs

AI Consumption

The AI Consumption cube provides a flexible workspace for analyzing AI usage and costs across your organization. You can build custom reports to understand AI spending, token consumption, pricing, usage trends, and other AI-related metrics using interactive tables and charts.

Use the AI Consumption cube to create reports that help answer questions such as:

  • How much are we spending on AI services?

  • Which models or providers generate the highest costs?

  • How many prompt and completion tokens are being consumed?

  • Which teams, users, or projects use AI the most?

  • How has AI usage changed over time?

The Analytics workspace lets you drag and drop fields to quickly build reports without writing queries.

Before using AI Consumption

Ensure that:

  • AI consumption data has been successfully ingested into Yarken.

  • The required AI dimensions and metrics are correctly mapped during ingesting AI Usage data.

Analytics workspace

The Analytics workspace uses a drag-and-drop interface to build reports.

Fields panel

The Fields panel contains every metric and attribute available for reporting.

Use the search box to quickly locate fields.

Fields are organised into categories such as:

  • AI Metrics

  • AI Attributes

  • Entities

  • Timeline

Simply drag a field into the required area of the Visualization panel.

AI metrics

Metrics are numerical values used for calculations and aggregations.

Common AI metrics include:

Metric

Description

Total Billed Cost

Total billed AI cost.

Total Effective Cost

Actual cost after discounts or negotiated pricing.

Total List Cost

Standard published pricing.

Total Contracted Cost

Cost based on contracted pricing.

Input Tokens

Total prompt tokens consumed.

Output Tokens

Total completion tokens generated.

Total Tokens

Combined prompt and completion tokens.

Cached Tokens

Tokens served from cache.

Consumed Quantity

Total AI consumption quantity.

Pricing Quantity

Quantity used for pricing calculations.

Note:

The available metrics may vary depending on your AI data source and configuration.

AI attributes

AI Attributes provide descriptive information about your AI consumption data. Unlike metrics, which measure costs or usage, attributes are used to group, filter, and analyze the data from different perspectives.

You can combine multiple attributes with AI metrics to answer questions such as:

  • Which users generated the highest AI costs?

  • Which AI models are used most frequently?

  • Which services or providers consume the most tokens?

  • How does AI usage vary by region, department, or billing account?

The AI Consumption cube includes attributes across several categories.

Supported AI attributes (expand to see the list)

General information

These attributes identify the source and timing of AI usage records.

Attribute

Description

Entity Code

The entity associated with the AI consumption record.

Loaded Date

The date the data was loaded into Yarken.

Usage Date

The date when the AI service was consumed.

Billing Period Start

Start date of the billing period.

Billing Period End

End date of the billing period.

Charge Period End

End date for the applicable charge period.

Ingested At

Timestamp when the record was ingested.

Month

Reporting month.

Year

Reporting year.

File Name

Source file used during ingestion.

Transaction ID

Unique identifier for the usage transaction.

Provider and service information

These attributes identify the AI platform and services generating the usage.

Attribute

Description

Provider

AI service provider (for example, OpenAI or Azure OpenAI).

Publisher

Organization or publisher offering the AI service.

Collector Source

Source from which the usage data was collected.

Invoice Issuer

Billing organization issuing the invoice.

Service

AI service being consumed.

Service Category

Category of the AI service.

Billing information

Use these attributes to analyse AI consumption by billing structure.

Attribute

Description

Billing Account ID

Identifier of the billing account.

Billing Account Name

Name of the billing account.

Sub Account ID

Identifier of the billing sub-account.

Sub Account Name

Name of the billing sub-account.

Billing Currency

Currency used for billing.

Pricing Unit

Unit used for pricing calculations.

Consumed Unit

Unit in which the service was consumed.

Resource information

These attributes identify the user or resource responsible for the AI consumption.

Attribute

Description

User

User associated with the AI request.

Resource ID

Unique identifier of the AI resource.

Resource Name

Name of the AI resource.

Region ID

Identifier of the deployment region.

Region Name

Geographic region where the service was hosted.

Usage classification

These attributes describe how the AI consumption is categorised.

Attribute

Description

Charge Category

Category assigned to the charge.

Charge Class

Classification of the charge.

Charge Frequency

Frequency of the charge.

Charge Description

Description provided for the charge.

Interaction Type

Type of AI interaction or request.

Pricing Category

Pricing model applied to the usage.

Model and SKU information

These attributes identify the AI model and pricing SKU associated with the consumption.

Attribute

Description

SKU ID

Unique SKU identifier.

SKU Price ID

Pricing identifier for the SKU.

SKU Meter

Meter used for billing.

Model

AI model used to process the request.

Model Version

Version of the AI model.

Modality

AI capability, such as text or image generation.

Pricing information

These attributes provide pricing details for the AI service.

Attribute

Description

List Unit Price

Published unit price before discounts.

Contracted Unit Price

Contracted unit price after negotiated pricing.

Commitment Discount ID

Identifier of the applied commitment discount.

Commitment Discount Name

Name of the commitment discount.

Commitment Discount Category

Category of the commitment discount.

Commitment Discount Status

Status of the applied commitment discount.

Tags and business dimensions

Tags allow AI usage to be grouped according to business-specific metadata.

Available tag attributes include:

  • Tags

  • Usage Pool

  • Application Tag

  • Product Tag

  • Service Tag

  • Team Tag

  • Department Tag

  • Environment Tag

  • Cost Center Tag

  • Customer Tag

  • Component Tag

  • Tech Stack Tag

These fields help you build reports by organisational, operational, or application-specific dimensions.

Financial dimensions

These attributes support cost allocation and financial reporting.

Attribute

Description

Account

Financial account associated with the charge.

Cost Center

Cost center responsible for the AI consumption.

Vendor

Vendor providing the AI service.

Entities

The Entities section allows reports to be analysed by organisational entities when multi-entity support is enabled.

This enables reporting across:

  • Business units

  • Departments

  • Companies

  • Legal entities

Timeline

Timeline fields are used to analyse AI usage over time.

Examples include:

  • Date

  • Month

  • Quarter

  • Year

These fields are commonly used in trend reports and time-series visualizations.

Example reports

The AI Consumption cube can be used to build reports such as:

  • AI spend by provider

  • AI spend by model

  • Daily AI cost trend

  • Prompt versus completion token usage

  • Total token consumption by user

  • AI usage by department

  • AI cost by project

  • Top AI consumers

  • Cached token utilisation

  • Effective cost versus billed cost

  • AI consumption by workspace

  • Monthly AI spending trend

Tips

  • Start with a Data Table to validate your data before switching to charts.

  • Apply filters early to improve report performance.

  • Use Timeline fields when building trend reports.

  • Combine multiple AI metrics to compare costs and token usage.

  • Save commonly used reports for quick access.

Related content