Paying for a full month of server time when your CI runners only work during business hours wastes money. Hourly billing at €0.0063/hr means an Ion runner that runs 8 hours a day costs about €0.05 for the day — against €4.50 for a full month. Spawn runners when your pipeline starts, tear them down when it finishes, and pay only for the compute you actually used from your prepaid credit.
Each KVM VPS deploys in under 60 seconds, and our Customer API (public beta) can order and cancel servers, making automated runner lifecycle management practical. Your CI orchestrator (Jenkins, GitLab CI, GitHub Actions, or a custom script) can provision a fresh Voxa VPS, run the build in a clean environment with no state contamination from previous runs, and destroy the instance when complete.
For always-on infrastructure — Kubernetes control planes, container registries, artifact stores, and monitoring stacks — monthly billing on the same KVM VPS keeps costs flat. vCPU and disk I/O are shared with the other virtual machines on the host, so latency-critical pieces such as a busy etcd cluster are better placed on a dedicated server.
Why Voxa for DevOps, CI/CD & Dev Environments
Pay for release weeks, not idle months
Pay for your actual build time. Ephemeral runners cost pennies per run, not euros per month.
Sub-60s provisioning
Runners are ready before your pipeline's first stage has time to time out. API-driven provisioning integrates with any orchestrator.
Clean KVM isolation per run
Each runner is a fresh isolated VM — no cross-build contamination, no shared filesystem risks, no state leaking between jobs.
Docker + Kubernetes native
Full root access means you install Docker, containerd, k3s, kubeadm, or any container runtime your workload needs.