What Is a Larger Runner in GitHub Actions?

A larger runner is a GitHub-hosted runner with more CPU, RAM, or a GPU than the standard two-core machines.

The default GitHub-hosted runners are modest, two CPUs and a few gigabytes of RAM. Larger runners offer beefier configurations for builds that are too slow or too memory-hungry for standard machines.

What it is

A larger runner is a configurable GitHub-hosted runner with more resources, for example 4 to 64 vCPUs, more memory, and optionally GPU or specific architectures. You define and label them in organization settings.

Using a larger runner

jobs:
  heavy-build:
    runs-on: ubuntu-latest-16-cores
    steps:
      - run: make -j16

How it works

After configuring a larger runner with a custom label, jobs select it via runs-on:. GitHub provisions the higher-spec machine and bills it at a higher per-minute rate than standard runners.

When to use them

Reach for larger runners when builds are CPU-bound and parallelizable, when memory pressure causes OOM failures, or when you need a GPU. For light jobs, standard runners are cheaper.

Why it matters

Larger runners can dramatically cut build time, but at a premium per minute. Managed runners like Latchkey offer comparable high-spec machines at lower cost, so you can size up without the steep bill.

Related concepts

Larger runners are selected by labels and have direct cost implications tied to Actions minutes.

Key takeaways

A larger runner has more CPU, RAM, or a GPU.
Select it with a custom label in runs-on:.
Faster builds but a higher per-minute cost.