GitHub Copilot Usage: Check Requests and Plan Your Work

GitHub Copilot reports usage as request counts, broken down by the type of interaction: completions, chat messages, and agent or cloud-agent actions. You can view these in your GitHub account settings under Billing and Plans, or, for organizations, in the organization's Copilot usage dashboard. Usage data tells you how much you are using the tool. It does not tell you which tasks are done, what is blocked, or where the next session should start. For that you need a separate record. fenbs keeps that record outside Copilot, so it is available when you switch tools or come back after a break.

Where to find your Copilot usage

For individual accounts: go to your GitHub profile, then Settings, then Billing and Plans. If you have Copilot enabled, a usage section shows your request activity. The exact layout changes as GitHub updates the UI, so check the GitHub Copilot documentation↗ (opens in a new tab) for the current path if the menu has moved.

For organizations: an administrator can see per-seat usage in the organization settings under Copilot. The dashboard shows which members are active and an approximate breakdown of how they are using the tool.

What the request categories mean

Copilot counts different types of requests differently. Inline completions are counted per suggestion. Chat messages count each exchange. Agent or cloud-agent actions (multi-step tasks that edit files or run commands) typically count as more than a single request because each step calls the model. Check GitHub's documentation for the exact counting rules for your plan, since they change as the product evolves.

• Inline completions: triggered as you type; each accepted or dismissed suggestion counts.
• Chat messages: each back-and-forth exchange in the chat panel.
• Cloud agent or Copilot agent actions: longer tasks that may involve multiple model calls and tool uses.
• Code review suggestions: Copilot's review comments on a pull request.

Usage limits and what happens when you reach them

Whether there are usage limits on your Copilot plan, and what they are, depends on the plan you are on. GitHub sets those limits and can change them. Check GitHub's current plan details rather than relying on any summary here. If you do hit a limit within a billing period, Copilot typically reduces functionality rather than stopping completely, though the exact behavior depends on the plan.

The difference between usage and task progress

Request counts tell you how much you used Copilot. They do not tell you which tasks are complete, which are blocked, or what the next step is. That distinction matters when you are handing work to another session or a different tool.

Say you have used a large number of agent requests today. That tells you the tool has been active. It does not tell anyone starting the next session what was accomplished, what was checked, or what is still open. If you save that information in fenbs before the session ends, the next session starts with a clear picture regardless of which assistant picks it up.

This separation matters for planning as well. If you know you need to continue a half-finished feature but your requests are limited for the session, you can save a handoff record now and continue in another supported tool. The task record in fenbs does not care which tool reads it.

Usage data is most useful as a baseline check rather than a performance metric. If completions are rarely accepted in a file type your team works in most, it is a signal that the context file is not set up well for that kind of code. If agent requests spike on a day when a large task was tackled, that is expected. Neither observation by itself tells you whether the work was done correctly. Pair the usage numbers with a task record that shows what was completed and verified, and you have a much fuller picture of what the tool actually produced.

Switching to another supported assistant

If your Copilot requests are limited and you need to continue working, fenbs connects to Claude Code, Cursor, Codex CLI, Gemini CLI, and GitHub Copilot. A handoff record in fenbs works with all of them. The work continues in a different tool without losing the context of what was done.

For an overview of the Copilot plans and what they include, see GitHub Copilot Business. For a comparison of alternatives, GitHub Copilot alternatives covers the options.

How to do this with fenbs

1. Create a free account and add a project for your current work.
2. Add tasks for each piece of work in progress. Note the status and the next step.
3. Connect GitHub Copilot from Integrations. Before each session ends, ask it to save task progress and Where we left off.
4. If Copilot usage limits affect your session, open fenbs to see the handoff record and continue in another connected tool.
5. The record of Decisions and rules and task progress is available to any connected assistant, so the choice of tool does not restart the explanation.