Let's talk GPUs. We reply within one business day.
Whether you need a custom bare-metal configuration, want to discuss enterprise pricing, or have a technical question, our team is ready to help.
- Sales
- sales@medjed.ai
Enterprise and bulk GPU reservations
- Support
- support@medjed.ai
Technical help and account issues
- Docs
- docs.medjed.ai
API reference and deployment guides
- Status
- status.medjed.ai
Live platform status
- X (Twitter)
- @medjedai
Product updates and benchmarks
- Blog
- blog.medjed.ai
Engineering notes and guides
Before you write
Looking for hourly GPU rates? The pricing page lists the Cloud KVM and bare metal SKUs currently published.
Need an account or want to provision an instance? Everything happens in the cloud console.
Found a security issue? Send it to support and mention the bug bounty program in the subject line.
What is GPU cloud rental service?
GPU cloud rental provides on-demand access to powerful GPU hardware over the internet. Our platform lets you rent GPUs from a global network, eliminating upfront hardware investment and maintenance costs. Ideal for AI/ML training, inference, rendering and other compute-intensive workloads.
How does GPU cloud rental work?
Pick a GPU based on your requirements (model, VRAM, CPU, RAM), rent it by the second or by the month, and access it via SSH, Jupyter notebooks or direct API integration.
What are the advantages of GPU cloud rental?
Cost savings versus buying hardware, flexibility to scale with workload, access to the latest GPU models without waiting for upgrades, global availability across regions, and simplified management through one dashboard.
What GPU models are available for rent?
We offer NVIDIA H100 NVL, H100 PCIe, H100 SXM, H200 and B200, plus RTX PRO 6000 and RTX 5090 bare metal servers. We regularly add the latest hardware as it becomes available.
What operating systems are supported?
Our KVM GPU instances run Linux, with Ubuntu images available by default. You can bring a custom image to install your own drivers and frameworks.
Can I run any workload on rented GPUs?
You can run a wide range of compute-intensive workloads including model training, inference, 3D rendering and scientific computing. Some workloads such as crypto mining are subject to provider policies.