Trending:
OpenAI admin controls GitHub Issues agent approvals Gemini in Chrome UK expansion Developer tools coverage
TECHSTAGED Better software. Smarter business.

Copilot usage metrics expand to OpenCode and more, enabling cross-tool visibility

By TechStaged

· 3 min read

Prepared with AI assistance from source material. Automated editorial review recorded . Automated checks can miss errors and do not represent human approval. Editorial policy and corrections.

Dashboard showing Copilot usage by client (OpenCode highlighted)
TechStaged-owned

Summary

  • The Copilot usage metrics API now reports model usage from more tools, including OpenCode.
  • A new totals_by_client array is available in the enterprise, organization, enterprise-user, and organization-user 1-day and 28-day reports.
  • Each totals_by_client entry includes client, request_count, session_count, token_usage, totals_by_model.
  • Unrecognized clients are grouped under other; clients identified by the integration ID they send.

GitHub's Copilot usage metrics API now reports model usage from more of the tools developers use with Copilot, including tools like OpenCode that weren’t previously captured in reports. This adds visibility across a broader set of integrations.

NEW DATA STRUCTURE: TOTALS_BY_CLIENT AND PER-MODEL METRICS

The Copilot usage metrics API adds a new totals_by_client array in enterprise, organization, enterprise-user, and organization-user reports for 1-day and 28-day views. Each entry represents one client and includes the tool name, request counts, session counts, token usage, and a per-model breakdown with a model_selection_mode.

SCOPE AND ACCESS: WHO CAN SEE THE DATA

Access is available to enterprise owners, billing managers, organization owners, and others with a compatible role and the Copilot metrics policy enabled. Unrecognized clients are grouped under other. Clients are identified by the integration ID they send, and the mapping is best-effort rather than guaranteed identity.

MODEL-LEVEL BREAKDOWN AND MODE OF SELECTION

Totals_by_model provides the same request and token metrics broken down by model, with a model_selection_mode that can indicate user_selected, auto, or product_selected. Certain tools report product_selected because the product chooses the model. Copilot code review models may be reported as unknown.

WHAT THIS MEANS FOR GOVERNANCE AND DECISION-MAKING

Counting requests at the server level closes gaps where editor telemetry was missing for tools outside the IDE, enabling comparisons of request and token volume across tools and supporting governance decisions. This helps determine which tools drive Copilot usage and where to invest in tooling and governance.

LIMITATIONS AND UNCERTAINTIES

The change is additive and does not replace existing fields. Totals_by_client is empty when there is no usage to report. The client identification is best-effort, and unrecognized clients are grouped under other. Some nuances, such as product-selected reporting for certain clients, are described in the documentation.

NEXT STEPS FOR TEAMS

Teams should review their Copilot usage metrics policy settings to enable visibility across tools, and plan governance decisions based on per-tool and per-model data now available in the 1-day and 28-day reports.

SOURCES

Developer Tools GitHub adds confidential comments to repository security advisories • 2 min read Developer Tools Xcode 27 runner image GA on GitHub Actions: arm64-only macOS runners with one-Xcode-version-per-image • 2 min read Developer Tools GitHub expands SecurityAdvisory GraphQL API with five new fields and two filters • 2 min read Developer Tools Open Source License Compliance GA for GitHub Enterprise Cloud signals enterprise-wide governance • 2 min read Developer Tools GA for Enterprise GitHub Apps: What the Evidence Shows Beyond the Announcement • 3 min read Developer Tools JetBrains GameDev Day 2026 Recordings: Cross-Engine AI Workflows and IDE-Integrated Debugging Emerge from the Evidence • 3 min read AI & Automation NVIDIA Minerva launches for end-to-end observability in AI factories • 2 min read AI & Automation NVIDIA cuDSS GA: What the evidence says beyond the announcement • 2 min read