GitHub Added Per-Model Token Breakdown to Copilot Usage Reports
GitHub Changelog·low signal
A 2026-08-11 changelog entry added per-model token breakdown to the GitHub Copilot usage report, exposing consumption by individual model rather than as a single aggregate. This lands the same day as the MAI-Code-1.1-Flash release and MAI-Code-1-Flash deprecation, which makes it directly useful for measuring the cost delta of that forced migration. For teams running multiple Copilot models it is the first native way to attribute spend without external instrumentation.