What you get

The same workflows, a smaller bill.

  • Setup

    Your workflows as they are

    Install the GitHub app on an organization or a personal account, pick its repositories, then set runs-on to one of our labels. Steps, actions and secrets stay as they are.

  • Image

    What ubuntu-24.04 has

    The tool cache that setup-python, setup-node and setup-java look for, Docker without sudo, service containers, passwordless sudo. We rebuild it every week from GitHub's.

  • Isolation

    One machine per job

    Each job gets its own virtual machine, created from the image and destroyed at the end. Nothing a job leaves behind reaches the next one.

  • Startup

    Jobs start in seconds

    Runners wait ready in Montréal. Billing starts when GitHub reports the job as started, so the time to boot a machine is never on your bill.

  • Caches

    Next to the runners

    actions/cache, the setup-* caches and buildx type=gha stay in Montréal without a workflow change. The Docker cache, per repository, keeps the layers and images of the last successful job that changed Docker.

  • Billing

    By the started minute

    Taken from your credit, prorated by vCPU. The console and the API show the minutes of each repository, and why a job waits.

In practice

One line in your workflow.

Connect GitHub from the console and choose the repositories the app may serve. Then replace ubuntu-latest with one of our labels: the job runs on 2, 4 or 8 vCPU, and the npm cache below is kept in Montréal.

The runners guide

yaml
name: CI
on: [push, pull_request]

jobs:
  test:
    runs-on: ffxf-2vcpu-ubuntu-2404
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with:
          node-version: 22
          cache: npm
      - run: npm ci
      - run: npm test

Sizes

Sizes and prices.

GitHub: published prices of its Linux x64 hosted runners, October 2026. Taxes included at FFxF.

Limits

What to know before you start.

FFxF · Montréal

Run your next job in Montréal.

Connect GitHub from the console, then change runs-on in one workflow to try it.

Support & discussions

Technical questions, incident reports, or infrastructure discussions, the team is reachable on Discord, Telegram, X, Instagram, Reddit, and IRC.