GitHub Copilot Usage API Now Tracks AI Credits Per User

GitHub adds granular consumption tracking to Copilot metrics

GitHub has expanded its Copilot usage metrics API to include per-user AI credit consumption data, giving enterprise administrators and organization owners visibility into how individual developers drive AI costs.

The new ai_credits_used field appears in both single-day and 28-day user-level reports at the enterprise and organization levels. The metric draws from the same consumption data that feeds GitHub's usage-based billing system, providing a consistent view of how AI credits are spent across developer activity.

What enterprises can do with the data

The per-user credit tracking enables three core use cases for organizations managing Copilot deployments:

First, administrators can now connect consumption directly to value by viewing AI credit usage alongside existing activity metrics in the same reports. This unified view shows which work patterns drive the highest AI costs.

Second, the data reveals how AI credit usage distributes across teams and departments. Organizations can identify where Copilot delivers the most value and where adoption may need support or adjustment.

Third, enterprises can monitor day-over-day consumption patterns to forecast AI credit requirements and inform budget planning as they prepare for or operate under usage-based billing models.

Implementation details

The ai_credits_used field represents an overall per-user total across all Copilot activity. GitHub notes the metric is not currently broken down by feature, model, or surface—it's an aggregate number.

Importantly, this field serves as a metrics signal for analyzing consumption, not as a billed total. Organizations should continue to reference their billing systems for invoicing purposes.

Access to these metrics is limited to enterprise administrators and organization owners who already have permissions to view Copilot usage metrics through the REST API. The data appears in the user-level reports specifically, not in aggregate organizational summaries.

Why it matters

As AI coding assistants shift from flat-rate to consumption-based pricing models, visibility into per-user costs becomes essential for enterprise planning. Without granular consumption data, organizations struggle to forecast budgets, optimize usage patterns, or demonstrate return on investment.

This update gives technical leaders the data foundation to make informed decisions about Copilot deployment strategies. Teams can identify power users who may benefit from additional training, spot underutilization that suggests adoption barriers, and build data-driven cases for expanding or adjusting AI tool investments.

For enterprises evaluating or operating under usage-based billing, the ability to track consumption trends before they hit invoices provides crucial budget predictability.

The enhancement was first reported by GitHub in their changelog. Organizations can access full documentation through the Copilot usage metrics API documentation or participate in discussions through GitHub Community.