Oblivus Cloud

Gain access to the most extensive range of high-end NVIDIA GPUs and cutting-edge CPUs in the industry.

Per-Minute Billing

Take advantage of our billing system, which charges you by the minute, enabling you to optimize your utilization of resources.

Lightning-Fast API

Program our infrastructure according to your needs by accessing our system through our API.

No Quotas

Say goodbye to the hassle of lengthy and complicated quota verifications. With Oblivus Cloud, you can enjoy a quota-free experience!

Secure & Reliable

Eliminate the risk of downtime and safeguard your work and user data by using our Tier 2-4 datacenters, which are certified with ISO/IEC 27001, 27017, 27018, and 27701.

What We Offer?

The leading cloud service provider for workloads that require GPU acceleration

Our main service is providing a vast range of NVIDIA GPUs. As a specialized cloud service, we offer computing resources that can handle the complexity of your workloads, and our infrastructure enables you to scale up easily.

Get Started


Accessing high-performance computing has never been easier!

1. Register

Sign up on Oblivus Cloud.

2. Top-Up

Add as little as $5 to your account using a credit card, bank transfer, or cryptocurrency.

3. Deploy

Set up your instance in just 30 seconds!

Deploy, Modify, Rescale

Our infrastructure is incredibly flexible and offers unparalleled customization options. Unlike other cloud service providers, you can modify each major component of your instance. You can scale up to thousands of GPUs and vCPUs within seconds.

We also allow you to switch between GPU and CPU instances as needed. If you need to keep costs low, you can deploy a CPU instance, but if you require more computing power, you can easily modify it to a GPU instance.

Instance Restrictions

State-of-the-art Infrastructure

We provide access to 13 enterprise-grade NVIDIA GPU SKUs and the latest CPUs in the industry, ensuring that we have the right solution to meet your computing needs.

Pricing Availability


Developed by innovators, for everyone


Launch new instances in as little as 30 seconds.


Deploy, modify, and scale up your instances directly with our API.

Modern Networking

Experience unlimited 10Gbps to 100Gbps network connectivity for each of your instances.


Tailor your cloud workload according to your needs.


Use only what you need, when you need it with on-demand access.


Get up to a 50% improvement in pricing with reserved instances.

Network Storage

Don't limit yourself to local storage - expand your storage capacity with network-based disks.


Pay only for the services you use, billed on a minute-by-minute basis. No hidden fees or undisclosed expenses.


Create an organization and utilize the account's funds with other members of your team.


Industry-leading transparent pricing

GPU Instances


NVIDIA Quadro RTX 4000

Exceptional performance at the most affordable price!

Available Stock: 1556

NVIDIA Tesla V100

Outstanding machine learning performance without breaking the bank!

Available Stock: 185

$0.27/hr - $5.17/hr

More: H100, A100, etc.

Cutting-edge GPUs designed to meet the demands of your workloads!

Available Stock: 5464

CPU Instances



Highest performance, latest architecture!

AMD EPYC Rome & Intel Xeon Scalable

Outstanding cost-effective performance!

Intel Xeon v4

General purpose with high core counts and boost clocks!

Intel Xeon v3

A great performance-to-cost ratio for the most extensive batch workloads!

Operating Systems & Images

Operating systems and images tailored to meet your workload requirements

Our platform offers a variety of operating systems for virtual machines and images for pods. Choose from CentOS, Ubuntu, Windows (BYOL), or bring your own image from Docker Hub, all at no extra cost. Furthermore, all our configurations include pre-installed NVIDIA drivers to support your workloads.


Designed specifically for artificial intelligence

Our custom operating system and image, OblivusAI, is built on Ubuntu 22.04 and includes all the necessary machine and deep learning libraries for your training needs. With OblivusAI, you can start training your models immediately without worrying about installing libraries.

Oblivus Cloud FAQ

When the instance is running, you will be billed for all of its components including vCPUs, RAM, storage, IP address, and possibly GPUs. However, when the instance is turned off, you will only be charged for the storage and the assigned IP address, which helps to keep costs low when the instance is not in use.

Our data centers are currently located in Chicago, New York and Las Vegas. If you need more information, please visit our data centers page.

Yes, every component of your instances, including GPU, CPU, RAM, storage, and IP addresses, is fully dedicated to you. We do not share any components with other customers, ensuring that you have the full resources available to you at all times.

Absolutely! We fully support rendering programs like Blender and others. You can perform any task you need on your instance without any limitations.

We offer two product lines: Oblivus Pods, which operates on bare-metal Kubernetes without a virtualization layer, and Oblivus Virtual Servers, which includes a virtualization layer.

The deployment time is contingent on factors such as system load, chosen operating system, and platform.

For Oblivus Virtual Servers, Linux-based operating systems can be deployed in approximately 30 seconds, while Windows virtual machines may take a few minutes for provisioning.

For Oblivus Pods, instances can be deployed in as little as 5 seconds if the image is available in our cache.

We offer complete KVM virtualization with root access and dedicated GPU pass-through for Oblivus Virtual Servers. Oblivus Pods is built on bare-metal Kubernetes.

Yes, you can launch multiple instances using our dashboard or API without any quota limitations..

Trusted by companies, institutions and researchers across the world

Looking for an affordable cloud solution that offers both flexibility and power?
Join us today!