Hardware
An overview of the hardware available to you.
Blaze for GitHub Actions provides you with a high-performance infrastructure to run your jobs on. With our hardware and software, you can run your jobs faster and more reliably than ever before.
Runner Tags
Blaze for GitHub Actions uses runner tags to specify hardware configurations. Runner tags determine the runner's operating system. These tags are used within the jobs.[job-name].runs-on
property in the workflow config. Please refer to the Run Your First Job guide for more information on how to use the runner tag.
¹ Concurrency units give you flexibility: you can run standard jobs (1 unit per job), use compute jobs for enhanced performance (2 units per job), or combine units for larger machines. Learn more about pricing
macOS
vCPUs | RAM | Runner Tag | Concurrency¹ |
---|---|---|---|
4 vCPU | 12 GB | blaze/macos-14 | 1x |
8 vCPU | 24 GB | blaze/macos-14:lg | 2x |
4 vCPU | 12 GB | blaze/macos-15 | 1x |
8 vCPU | 24 GB | blaze/macos-15:lg | 2x |
- The latest stable macOS can also be specified by using the blaze/macos-latest runner tag.
Linux
Linux runner subscriptions are hardware agnostic. You can use either ARM64 or x86_64 hardware with your Linux runner subscription. The latest stable Linux OS can also be specified by using the blaze/ubuntu-latest runner tag.
ARM64
Standard Runners
vCPUs | RAM | Runner Tag | Concurrency¹ |
---|---|---|---|
4 vCPU | 12 GB | blaze/ubuntu-22.04 | 1x |
8 vCPU | 24 GB | blaze/ubuntu-22.04:lg | 2x |
16 vCPU | 48 GB | blaze/ubuntu-22.04:xl | 4x |
32 vCPU | 96 GB | blaze/ubuntu-22.04:2xl | 8x |
Compute Runners
Compute runners feature enhanced CPU performance and high-speed memory while maintaining the same memory capacity as standard runners.
vCPUs | RAM | Runner Tag | Concurrency¹ |
---|---|---|---|
4 vCPU | 12 GB | blaze/compute/ubuntu-22.04 | 2x |
8 vCPU | 24 GB | blaze/compute/ubuntu-22.04:lg | 4x |
16 vCPU | 48 GB | blaze/compute/ubuntu-22.04:xl | 8x |
32 vCPU | 96 GB | blaze/compute/ubuntu-22.04:2xl | 16x |
x86_64
Standard Runners
vCPUs | RAM | Runner Tag | Concurrency¹ |
---|---|---|---|
4 vCPU | 16 GB | blaze/ubuntu-22.04-amd64 | 1x |
8 vCPU | 32 GB | blaze/ubuntu-22.04-amd64:lg | 2x |
16 vCPU | 64 GB | blaze/ubuntu-22.04-amd64:xl | 4x |
Compute Runners
Compute runners feature enhanced CPU performance and high-speed memory while maintaining the same memory capacity as standard runners.
vCPUs | RAM | Runner Tag | Concurrency¹ |
---|---|---|---|
4 vCPU | 16 GB | blaze/compute/ubuntu-22.04-amd64 | 2x |
8 vCPU | 32 GB | blaze/compute/ubuntu-22.04-amd64:lg | 4x |
16 vCPU | 64 GB | blaze/compute/ubuntu-22.04-amd64:xl | 8x |
Compute tier runners consume twice the concurrency slots of standard runners while delivering enhanced performance. The latest stable Linux OS can also be specified by using the blaze/compute/ubuntu-latest runner tag.
GPU
All macOS runners have a virtual Metal GPU available, allowing you to run some basic GPU tasks and tests. If you need a more powerful GPU, please contact us at contact@runblaze.dev.
=========
Runner configurations and hardware are subject to change on occasion as newer hardware becomes available. We strive to always maintain or improve performance levels.
On This Page