GPU & AI
AI Model Gateway
200+ models · one API key · prepaid credit
From
$10.00 credit pack
View plan
Save 25%
GPU & AI
GPU Cloud — H100
80 GB HBM3 · 1 GPU · per day
From
$29.90
/day
$39.90
● Available now SKU GT-GA-002
- ✓ No setup fee · cancel anytime
- ✓ 24/7 engineer support
- ✓ 99.99% uptime SLA
- ✓ USD billing, single invoice
Specifications
- GPU
- NVIDIA H100 SXM, 80 GB HBM3
- vCPU
- 32 threads
- Memory
- 256 GB DDR4
- Storage
- 1 TB NVMe
- Network
- 10 Gbps
- Regions
- Singapore, Denver
- Interconnect
- NVLink-ready
- Images
- CUDA 12, PyTorch, vLLM, TensorRT
Overview
Enterprise-grade compute for large-model training, high-throughput inference and simulation. An H100 SXM instance ships with 80 GB of HBM3, 32 vCPUs, 256 GB of RAM and 1 TB of NVMe storage over an NVLink-ready fabric.
Ideal for teams that need serious compute for a sprint without buying hardware or signing a multi-year reservation.
What's included
- Per-day billing, no commitment
- NVLink-ready fabric for multi-GPU jobs
- High-speed 10 Gbps network
- Preinstalled vLLM and TensorRT
- Persistent volumes available
- Dedicated GPU engineer on call
Common questions
How quickly can I get started?
Cloud VPS and GPU instances are provisioned automatically after signup. Our team confirms your configuration and credentials within one business day for managed services.
How is billing handled?
All prices are in USD. Monthly plans renew monthly; GPU instances are billed per day. We invoice by card or wire transfer and cancel any time before the next cycle.
Can I change plans later?
Yes. You can resize VPS plans from the console, and GPU or managed plans can be adjusted with your account engineer. There is no migration fee between our own plans.
Do you offer migration help?
Yes — send us your current stack and target region through the contact form and we will outline a migration path, usually at no extra cost for standard workloads.
Related plans
GPU & AI
