Skip to main content
Compute allows you to create and manage virtual machines (VMs). These VMs function similarly to your local machine but are hosted in the cloud. You can connect to them and use their GPUs and other computing resources in your ML/AI workloads. In the web console, Compute virtual machines are grouped into the following tabs on the Virtual machines page:
  • Standalone VMs
  • Kubernetes nodes
  • Container VMs
  • Job and endpoint VMs
    Only VMs for jobs and endpoints created before July 28, 2026, are shown. For jobs and endpoints created after this date, Serverless AI provisions and manages the underlying Compute resources.
The service is available in all Nebius AI Cloud regions. InfiniBand and InfiniBand Trade Association are registered trademarks of the InfiniBand Trade Association.

Getting started

Create your first set of VMs and connect to them

Types of virtual machines

Learn about VM configurations

Creating virtual machines

Learn how to create different types of VMs

Connecting to virtual machines

Learn how to connect to VMs using SSH and allow other users to access VMs
Create a GPU cluster and add VMs to it to interconnect them with InfiniBand

Testing InfiniBand connection

Learn how to run tests to check the InfiniBand network state and performance

Types of storage volumes

Choose a storage option for your VMs that fits your goals best

Managing volumes

Manage and configure disks and shared filesystems

Attaching and mounting volumes to VMs

Use volumes that you created on VMs

Monitoring VMs

Control GPU and vCPU states of your VMs

Monitoring volumes

Control disks and filesystems states