GPU INSTANCES

On-demand GPU compute built for AI

Launch top-tier GPU instances in seconds - preconfigured for ML, pay-by-the-minute, no commitment.

GPU INSTANCE PRICING

Our full range of high-end NVIDIA GPUs

GPUs	# of GPUs	VRAM/GPU	vCPUs	RAM (GB)	Storage (GB)	Price (USD)
NVIDIA H200	1x	141	24	240	3,000	3.50/h
NVIDIA H100 SXM	1x	80	24	240	3,000	2.90/h
NVIDIA H100 PCIE	1x	80	30	380	3,840	2.90/h
NVIDIA A100	1x	80	15	180	300	2.74/h
NVIDIA L40S	1x	48	15	90	1,600	1.55/h
NVIDIA V100S	1x	32	15	45	300	0.95/h
NVIDIA L4	1x	24	22	90	1,600	0.93/h
NVIDIA V100	1x	16	8	45	300	0.83/h

HOW IT WORKS

Flexible
GPU options
Top-tier GPUs at best-in-industry rates, from fractional to multi-GPU instances.
Optimized
for AI
Pre-installed OS, ML frameworks and drivers— tuned for instant productivity.
Suspend, resume, scale
Pause idle workloads to cut costs, resume on demand, and scale up when needed.
Best-in-industry rates
Transparent pricing. No ingress or egress fees — or hidden fees of any kind. Ever.

WHY ORI GPU INSTANCES?

Optimized to help you meet ambitious timelines and budgets with ease

SAVINGS
81%
Cost savings, compared to other cloud hyperscalers
FLEXIBILITY
40+
Available GPU configurations
ACCESS
15+
Regions of service across the globe

WHY ORI GPU INSTANCES

From idea to launch

ITERATIVE INNOVATION

Built for fast, iterative development

Purpose-built for fast, iterative cycles—preprocessing, training, fine-tuning, and experiment at scale

EASE OF INTEGRATION

Launch from CLI, UI, or API

Launch and scale via CLI, UI, or API—whichever fits your stack.

Private Cloud

lets you build enterprise AI flexibly and in control

Why developers love Ori

Vinay Maniam
Founding Engineer, nCompass

Chart your own
AI reality