GitHub Copilot code review now consumes Actions minutes on private repos
GitHub Copilot code review starts consuming GitHub Actions minutes on June 1, 2026 for private repositories, on top of Copilot AI Credits. Teams using automatic PR reviews should check budgets, runner settings, and review frequency before the new usage pattern surprises billing owners.
Starting June 1, 2026, GitHub Copilot code review consumes GitHub Actions minutes when reviews run on private repositories using GitHub-hosted runners. GitHub says Copilot usage is also billed as AI Credits under the usage-based model, so teams may see both Copilot and Actions usage affected by automated reviews. Public repositories are not changed in the same way because Actions minutes remain free there.
Key takeaways
- The change applies to Copilot Pro, Pro+, Business, and Enterprise plans, according to GitHub's changelog.
- Private-repo Copilot code reviews now consume Actions minutes when they run on GitHub-hosted runners.
- Copilot code review usage also counts under the AI Credits model, so this is a usage-budget issue, not only a Copilot feature toggle.
- GitHub says self-hosted runners and larger GitHub-hosted runners are supported, but billing behavior differs by runner type.
- Teams with automatic Copilot reviews should check whether every pull request now creates avoidable Actions spend.
Practical LinkLoot angle
This is a workflow-cost story for engineering leads, not just a pricing footnote. Automatic AI reviews can be useful, but the cost profile changes when every private pull request may spend both model credits and CI minutes. The practical move is to decide where Copilot review is mandatory, where it should be manually requested, and where normal tests plus human review are enough.
| Review setup | Best use | Cost or control issue | What to check |
|---|---|---|---|
| Automatic Copilot review on every private PR | High-risk repos where every change needs an AI pass | May create repeated Actions-minute usage across routine PRs | Review frequency, budget alerts, runner type |
| Manual Copilot review | Targeted use on complex diffs or unfamiliar code | Depends on developer judgment and review habits | Team policy and PR templates |
| Public repository review | Open-source workflows where Actions minutes remain free | Still uses Copilot AI Credits where applicable | Copilot quota and model multipliers |
| Self-hosted runner path | Teams that already operate controlled runners | Operational overhead shifts back to the team | Runner security, isolation, maintenance |
For LinkLoot readers, this belongs in the same bucket as agentic workflow automation: the agent is only useful if its trigger, budget, and approval path are explicit. A good policy is simple: enable automatic AI review only where the failure cost is higher than the review cost, then use manual review requests for everything else.
What to verify before you act
First, verify whether Copilot code review is enabled automatically at the organization, enterprise, or repository level. Then check Actions minute consumption on private repositories after June 1, not just Copilot usage dashboards. If billing owners and engineering leads are different people, send the changelog to both groups because this change crosses those boundaries.
Octomind's analysis gives example minute ranges for medium and large reviews, but those estimates are not GitHub's billing guarantee. Treat them as a planning prompt, then measure your own repos: repository size, diff size, runner choice, and review frequency will decide the real cost.
Source check
GitHub's changelog confirms the June 1, 2026 effective date, the private-repository Actions-minute change, the affected Copilot plans, and the recommendation to review billing, usage, budgets, and runner settings. GitHub Docs repeats the June 1 notice inside Copilot code review documentation. Octomind independently discusses the cost impact and gives practical examples for teams that run automatic reviews.
For a broader automation policy checklist, see LinkLoot's guide to AI workflow automation.
Yes. Starting June 1, 2026, Copilot code review runs consume GitHub Actions minutes on private repositories when they run on GitHub-hosted runners.