Copilot usage metrics API now includes client‑level totals

The latest update to the Copilot usage metrics API introduces a totals_by_client array in the 1‑day and 28‑day enterprise, organization, enterprise‑user, and organization‑user reports. The array records model requests from a broader set of client tools—including OpenCode, VS Code, JetBrains, and others—so you can now see which external tools are driving Copilot traffic.

New client‑level totals in the API

Each entry in totals_by_client identifies a client (e.g., opencode, vscode, jetbrains) and provides:

• request_count: successful model calls from that client.
• session_count: distinct sessions observed.
• token_usage: prompt tokens, output tokens, and average tokens per request.
• totals_by_model: the same metrics broken out by model, with a model_selection_mode indicating whether the model was user‑selected, auto‑chosen, or product‑selected. Unrecognized clients appear under other, and unknown selection modes are labeled unknown.

When no usage exists, the array is empty, preserving backward compatibility.

Operational impact

Previously, usage from tools that did not emit editor telemetry was omitted from totals_by_ide. By counting requests on the server side, you now have a complete picture of model traffic across all integrations. This data supports:

• Capacity forecasting based on request and token volume per client.
• Tool‑specific governance decisions, such as restricting or promoting certain integrations.
• More accurate cost attribution for enterprise billing.

Implementation considerations

Clients are identified by the integration ID they send with each request, which is a best‑effort grouping rather than a verified identity. request_count reflects model calls, not individual prompts, so it should not be directly compared with prompt‑level metrics. The new fields are additive; existing report structures remain unchanged.

Security and compliance notes

Access to the enhanced reports is limited to enterprise owners, billing managers, organization owners, and any custom role granted the View Copilot Metrics permission. The Copilot usage metrics policy must be enabled for the data to appear. Because client identification is not guaranteed, treat the grouping as advisory when designing compliance or audit processes.

Related CloudNinjas coverage: AI engineering.

What This Means For Practitioners

Enable the Copilot usage metrics policy if it is not already active, and verify that the appropriate roles have the View Copilot Metrics permission. Update monitoring dashboards to ingest totals_by_client and correlate request counts with token usage for each tool. Use the new visibility to evaluate whether any external integrations require additional governance, capacity adjustments, or cost‑control measures.