Anex lets you run GPU workloads on cheap cloud GPU providers (Vast.AI, RunPod) while managing them through your existing Kubernetes cluster. You write standard Kubernetes manifests, and Anex takes care of renting machines, setting up networking, and making remote GPU pods feel like they’re part of your cluster.
Native Kubernetes
Standard Pods, Deployments and Jobs scheduled to a Virtual Kubelet node. No new APIs to learn — just pod annotations to pick the machine.
Open Source
Source on github.com/anex-sh/anex. Pluggable cloud providers; Vast.AI and RunPod supported today.
Cheap GPUs
Rent only what you need, when you need it. Pods scale up by provisioning machines on demand and tear them down when removed.