GitHub Copilot Usage Report Now Shows Per-Model Token Breakdown
GitHub Copilot's usage report now provides a detailed breakdown of tokens consumed per AI model, including input, output, cache read, and cache write tokens. This enhancement directly links token usage to AI credit consumption, helping users understand charges and optimize costs. The new visibility allows for better communication of usage details to stakeholders. This feature is available to admins of Copilot Business and Enterprise, and all users of Copilot for Individuals, accessible via the AI usage report in billing settings.
Features (1) ›
- Per-Model Token Breakdown in AI Usage Report
The GitHub Copilot usage report now shows a detailed breakdown of input, output, cache read, and cache write tokens for each AI model. This allows users to understand how AI credits are consumed, explain charges, and identify opportunities for cost reduction.
https://github.blog/changelog/2026-08-11-per-model-token-breakdown-in-the-usage-report
Related releases
- GitHub Copilot introduces Agent Plugins 1.0 for cross-client development GitHub Changelog ·
- GitHub Details Organization-Level Repository Rule Insights GitHub Changelog ·
- GitHub Enterprise Server 3.22 Release Candidate Now Available GitHub Changelog ·
- GitHub Copilot for JetBrains adds persistent memory, Ollama integration, and enterprise controls GitHub Changelog ·
- Automatically Migrate Branch Protection Rules to Repository Rulesets GitHub Changelog ·
- GitHub Copilot's MAI-Code-1-Flash model slated for deprecation GitHub Changelog ·