Depot GitHub Actions Runners: Setup Guide (2026)
Depot GitHub Actions runners are a drop-in replacement for GitHub-hosted runners: per Depot's docs, you adopt them by changing the runs-on label. Here is an honest getting-started walkthrough.
Depot provides managed GitHub Actions runners you attach to existing workflows without rewriting them. According to Depot's docs at depot.dev/docs/github-actions/overview, adoption is a runs-on label change, and the integrated distributed cache works with no configuration changes. This guide covers the general setup path and the choices you make along the way. It is written to be accurate and vendor-neutral; verify exact steps and any account limits in Depot's own docs, since product UIs change.
1. Create a Depot account and connect GitHub
Sign up at depot.dev. Depot offers a 7-day free trial with no credit card per its pricing page. Connect the Depot GitHub App to the GitHub organization (or organizations) whose workflows you want to run on Depot runners. Runner usage is scoped to a Depot project and organization, so make sure the account, GitHub org, and project are linked before you expect jobs to be picked up.
2. Choose a runner type
Depot offers Intel x86, ARM, macOS, Windows, and GPU runners. Per Depot's pricing, GPU runners require the Business plan. Pick the type that matches your build: x86 for most jobs, ARM for arm64 targets, macOS/Windows for those platforms, GPU for accelerated workloads. Each type maps to a specific runs-on label from Depot's docs.
3. Swap the runs-on label
The core change is one line. Replace runs-on: ubuntu-latest (or your current label) with the Depot runner label from Depot's docs for the type and size you chose. Your steps, actions, secrets, and caching keys do not change. That is the whole point of a drop-in runner: only the runs-on line moves.
4. Let the cache warm up
Depot's integrated distributed cache is designed to work with no config changes, with upload/download up to ~1000 MiB/s per Depot. The first few runs populate the cache, so expect the biggest speedups to appear once it is warm rather than on the very first job. If you also use Depot for Docker image builds, its BuildKit builders sit next to the runners so build layers benefit too.
5. Watch minutes and plan limits
Depot runners are billed per minute (tracked per second, with no enforced one-minute minimum per Depot). Depot's Developer plan includes 2,000 minutes then $0.004/min, and the Startup plan includes 20,000 minutes; Depot states runners run at about half the cost of GitHub-hosted. During the 7-day trial, keep an eye on included minutes so a large pipeline does not exhaust them mid-evaluation. Verify current numbers at depot.dev/pricing.
Or try self-healing managed runners
If you reach setup and realize your real problem is flaky re-runs rather than raw speed, it is worth a look at Latchkey. It is also a drop-in managed runner reached by a runs-on swap, but it adds self-healing CI that detects, fixes, and retries transient and mechanical failures automatically, alongside up to 58% lower per-minute cost than GitHub Actions. You can pilot it on one workflow next to Depot and compare.