CI Runners
GitHub Actions on machines that are yours alone
Run your existing workflows on Quant-hosted runners in Australia. Each pool has its own machines, native arm64 and x64, and a fixed monthly price.
Isolation
No shared machines
A runner pool is a set of virtual machines that only your organisation uses. Each job starts in a fresh container on a single-use runner, and the runner is removed when the job ends.
Two customers never share a machine, a Docker daemon, a disk or a cache
Jobs from public repositories are refused, so a fork cannot run code on your pool
The GitHub App reads job events and registers runners. It has no access to your code
- No personal access tokens to create or rotate
How a job runs
| Tier | Concurrent jobs | Each job gets |
|---|---|---|
| Small | 2 | 2 vCPU, 7 GB memory, 50 GB disk |
| Medium | 4 | 4 vCPU, 14 GB memory, 75 GB disk |
| Large | 8 | 4 vCPU, 14 GB memory, 75 GB disk |
The tier is a hard cap. When every slot is busy, new jobs wait in the queue and start when a slot is free.
Secrets and variables
Keep CI secrets off GitHub
Store secrets and variables in Quant and read them with the syntax your workflows already use. A job receives only the values for its own repository and branch, when it starts on your pool.
Scope a value to the whole organisation or to one repository
Limit deploy credentials to named branches, such as main
- Jobs for pull requests from forks receive no values
If the values cannot be delivered, the job fails before its first step
GitHub OpenID Connect works, so cloud deploys need no stored access keys
Secrets and variables documentation
Running in three steps
You keep your repositories, your workflows and your logs on GitHub.
Create a pool
Choose a name, a tier and the architectures in the Quant dashboard. The pool is ready in a few minutes.
Cloud → CI runners → Create poolConnect GitHub
Install the Quant CI runners app on your GitHub organisation and choose the repositories that may use the pool.
All repositories, or the ones you selectChange runs-on
Point a job at the pool. You can move one job at a time and leave the rest on GitHub.
runs-on: [self-hosted, quant, build, x64]What teams use it for
The same workflows, on machines you do not share.
Native ARM64 image builds
Build images for ARM64 environments on Graviton runners. A native build is much faster than a build under emulation.
- arm64 and x64 in one pool
- Docker and Buildx ready
- Multi-architecture matrix builds
Builds that stay in Australia
Source code, build output and secrets are processed on machines in Melbourne, on the platform that hosts your applications.
- Jobs run in Melbourne
- Isolated machines per customer
- Available on QuantGov
Deploys to Quant Cloud
Build and push your application image next to the registry and the environments it deploys to.
- Quant Cloud GitHub Actions
- Docker Compose support
- Service containers
Deploy credentials for one branch
Give a deploy key only to jobs on your release branch. Pull request jobs never receive it.
- Branch lists per value
- Repository scope
- Masked in job logs
A CI bill you can predict
A pool has a fixed monthly price. A busy month costs the same as a quiet one, and the dashboard tells you when a larger tier would help.
- No per-minute charges
- Usage and queue times per pool
- Upsize at any time
Gradual migration
Move one job, check it, then move the next. Jobs you have not changed keep running on GitHub-hosted runners.
- No workflow rewrite
- Same secrets syntax
- Logs stay in GitHub
Run your next build on your own pool
CI runners are available to every organisation with Quant Cloud access on a Teams Lite plan or higher.