Instant access to NVIDIA H100, L40S and V100 GPUs

Get immediate access to up to 8 NVIDIA® GPUs, along with CPU resources, storage and additional services through our user-friendly self-service console.

Up to 8 GPUs on demand in 5 minutes

The self-service console allows you to quickly set up an account, create and start virtual machines on demand, and stop them once the load decreases.

Scale your resources up and down

The console provides access to a single GPU host, with the flexibility to scale resources through our support team as needed.

Only $25 to start

A minimum payment of $25 will be charged and deposited to your account once you submit your bank card details. Bank transfer or direct debit are also available.

Get in touch with 24/7 support

You can contact our technical support through the support center in the console. Our support engineers are key members of Nebius AI team and ready to assist you with any issues.

Step-by-step guides

Our comprehensive documentation is designed to help you navigate Nebius AI services effortlessly. Use the detailed guides for setup and configuration, API references, best practices, FAQs and tutorials.

Request new features

Your feedback helps us improve the platform and better meet your needs. If you have a feature request or suggestion, submit your ideas. We promise they will reach our development team.

Nebius AI console walkthrough

Watch this video for a quick overview of our console. Khamset demonstrates how to set up an account, launches a VM, walks you through our available services and shows how to monitor your infrastructure and grant granular access to resources.

H100 SXM5

Platform: Intel® Sapphire Rapids

  • vRAM: 80 GB
  • RAM: 160 GB
  • Number of vCPUs: 20
  • Max GPUs per VM: 8
  • Price: $4.85 GPU/h


Platform: Intel Ice Lake

  • vRAM: 48 GB
  • RAM: from 32 to 384 GB
  • vCPUs: from 8 to 96
  • Max GPUs per VM: 1*
  • Price: $1.55 GPU/h

V100 SXM

Platform: Intel Cascade Lake

  • vRAM: 32 GB
  • RAM: 48 GB
  • Number of vCPUs: 8
  • Max GPUs per VM: 8
  • Price: $0.80 GPU/h

V100 PCIe

Platform: Intel Broadwell

  • vRAM: 32 GB
  • RAM: 48 GB
  • Number of vCPUs: 4
  • Max GPUs per VM: 8
  • Price: $0.73 GPU/h


Which payment methods do you accept?

We accept debit or credit card payments and bank transfers.

Start training and scaling your ML model today

4 GPU per VM coming this summer