Track AI spend in DX: Meet the new AI cost management report
Kali Watkins
Product Marketing
Organizations are increasingly using multiple AI tools across different teams, each with its own dashboard, pricing model, and usage data. As adoption grows, it becomes harder to understand AI investment across the organization, identify where costs are concentrated, and evaluate spend alongside engineering output.
To help solve this, we recently launched AI cost management in DX: a new report that brings together AI spend and token usage across tools and ties it to engineering output.
The report unifies spend and token data across AI tools, ties it to your team structure, and connects it to metrics such as estimated cost per PR. Teams can use it to forecast AI costs and token usage, analyze spend at the team and contributor level, and identify trends in efficiency over time.
How the report works
AI cost management combines AI tool metrics with source code management data and, when enabled, AI Code Insights session data.
The report is designed to help engineering leaders understand where AI investment is going, how it is changing over time, and whether spending aligns with engineering output. Teams can use it to:
-
Understand AI spend across tools. Bring together spend and token usage data from multiple AI coding tools into a unified view. Teams can track trends over time and view breakdowns of input, output, and cached token usage.
-
Analyze AI investment by team and contributor. Break down spend and usage by team, contributor, attribute, or AI tool to identify where costs are concentrated across the organization.
-
Connect AI cost to engineering output. Compare estimated AI spend with merged PR volume to calculate an estimated cost per PR. Teams can use this metric to identify differences in efficiency across teams and investigate whether factors such as model selection, tool usage patterns, or workflow differences are contributing to higher costs.
-
Forecast future usage and cost. Project future spend and token usage based on historical trends to support budgeting and planning.
-
Drill into AI-assisted development activity. Explore contributor-level AI coding sessions and repositories where AI-assisted code is landing. (Available for teams who have enabled AI Code Insights)
The report supports a growing set of out-of-the-box AI tool integrations, including Claude and Cursor, and can also ingest custom AI cost data through the aiToolMetrics.push API. GitHub Copilot usage-based billing data is coming soon.
Getting started
AI cost management is available now in DX for organizations using Data Cloud with at least one supported AI tool connector.
To learn more, contact your DX account rep or read our docs.