Runner Concurrency - CI/CD Glossary Definition
Runner concurrency is the number of jobs that can execute in parallel across your runner fleet at one instant. It is capped by the number of available runners (and, on GitHub-hosted plans, by an account-level concurrency limit on how many jobs may run simultaneously).
The bottleneck
If 50 jobs are dispatched but only 20 runners are free, 30 jobs queue. Raising runner concurrency means provisioning more runners or a higher hosted-concurrency tier. The trade is cost: idle runners waiting for the next burst still cost money unless they scale to zero.
Hosted limits
GitHub-hosted concurrency is tied to your plan (for example 20 concurrent jobs on GitHub Free, more on Team and Enterprise). Self-hosted and managed runners let you set concurrency by simply running more machines.