Free tool

What do your CI runners cost?

Enter your GitHub Actions minutes a month by runner, or paste your usage report. See the monthly bill, what smaller runners and caching would save, the same work on GitLab and CircleCI, and when self-hosted runners pay off.

  • Prices come from each vendor's own pricing page, with the date we checked them.
  • Every result shows its formula and inputs, so you can check the math.
  • Everything is calculated in your browser. Nothing you enter is sent anywhere.

Loading prices…

Your GitHub Actions use

Its included compute is taken off the bill.
Your estimate: dependency and Docker layer caching, cancelled superseded runs, skipped docs-only runs.

Paste your GitHub usage report (CSV)

Download the summarized or detailed usage report from the Metered usage page of your billing settings and paste the CSV. It replaces the runners below. Only GitHub Actions rows measured in minutes are read.

Days. The report's minutes are scaled from the days it covers to this.

    Minutes a month by runner

    To test a smaller runner, pick it under "Move to" and enter how long the same jobs take there, as a % of today's time (160 means 60% longer).

    RunnerMinutes a monthMove toRun time there (%)Remove

    Self-hosted runners (your costs)

    GitHub doesn't charge per minute for self-hosted runners, so their cost is what you pay to run them. Enter your own numbers; leave a field empty and the break-even waits for it.

    From your cloud's pricing page or bill.

    730 is always on. Fewer with autoscaling.

    Patching images, scaling, fixing broken runners.

    Storage, egress, NAT, a cache server.

    Compare with
    Seat prices are not included, only compute minutes.
    CircleCI User charges are not included, only compute credits.

    How to check

    Find your own numbers

    • Download the usage report from your organization's billing page and rank Actions minutes by repository, workflow and runner type. In the top workflows, look for long runs, superseded runs and docs-only full runs.
    • Key dependency caches on the lockfile hash, with actions/cache and hashFiles('**/package-lock.json') or the cache option on setup-* actions. For Docker, use cache-from: type=gha and cache-to: type=gha,mode=max.
    • Cancel superseded runs on pull request workflows only, with a concurrency group and cancel-in-progress: true. Never on main, release or deploy workflows: it can cancel a deploy midway.
    • Size runners by measurement: run the same workflow on both sizes, and switch only if run time drops at least in proportion to the per-minute price.
    • The full walk-through is gotcha 10 in our free Cloud Bill Gotchas guide.

    Privacy

    Your numbers stay on this page

    • The tool runs in your browser. The page loads its own price list from this site and sends nothing you type or paste.
    • Its Content-Security-Policy only allows connections back to this site, and it loads no third-party scripts, fonts or trackers.
    • A shared link keeps your inputs after the #. Browsers never send that part of a link to a server.
    • The only request with something you typed is the optional email signup. It sends your email address and the tool's name, nothing else.

    Want us to look at the real thing?

    Book a free 30-minute consult: we'll look at your bill with you, no obligation and no sales pitch. Or check your own cloud bill first with our free Cloud Bill Gotchas Checker.

    Book a free 30-minute consult