// NVIDIA H100

Rent NVIDIA H100 GPUs
On-demand, in India.

H100 SXM5 with 80GB HBM3 — our most-deployed GPU. INR-native billing, sub-2-minute provisioning, GST invoices. No dollar conversion, no long-term contract.

Architecture
Hopper
VRAM
80GB HBM3
RAM
200GB
vCPUs
16
Starting at
₹225/hr

99.9% SLA Uptime

Backed by contract. Auto credits if we miss it.

Per-Minute Billing

Pause anytime. Billing stops instantly. Resume same state.

< 2 Min Provisioning

GPU live and SSH-ready in under 2 minutes. No queues.

// Flexible Compute

Pick your instance.

From instant Jupyter workbenches to full root-access VMs. Switch anytime.

H100 SXM5

Hopper
E3AvailableMOST DEPLOYED
VRAM80GB HBM3
RAM200GB
vCPU16
225/hr

A100 SXM4

Ampere
E2Available
VRAM80GB HBM2e
RAM160GB
vCPU16
125/hr

RTX 4090

Ada
CAvailableBEST VALUE
VRAM24GB GDDR6X
RAM64GB
vCPU8
50/hr

L40S

Ada
Available
VRAM48GB GDDR6
RAM128GB
vCPU16
113/hr

A100 PCIe

Ampere
E2Available
VRAM40GB HBM2e
RAM112GB
vCPU16
120/hr

RTX 3090

Ampere
CAvailable
VRAM24GB GDDR6X
RAM48GB
vCPU8
37/hr
// What's Included

Compare environments.

A full breakdown of capabilities across every compute environment.

Feature
Workbench
Bare Metal
Reserved
INR Billing
JupyterLab
VS Code Server
SSH Access
Full root access
Custom Docker images
Persistent storage
Multi-GPU (8x)
Pre-built ML stacks
GST Invoice
99.9% SLA
Dedicated support

Frequently asked questions

The H100 (Hopper architecture) is built for large-scale LLM training, fine-tuning, and high-throughput inference. Its 80GB of HBM3 memory and high memory bandwidth make it the standard choice for models that won't fit comfortably on an A100 or RTX 4090.

// Ready to deploy

Your first GPU in
under 2 minutes.

SSH in under 1 minute · INR billing · GST invoices included

Talk to Sales