Hyperstack review

Facts from Hyperstack’s own pricing and docs pages, September 202675 read from a quoted sentence on the page we archived.

Summary

Hyperstack rents GPUs from $0.15 per GPU-hour on demand, bills per minute, and asks no minimum commitment. Egress is free. Hyperstack does not publish a refund policy.

SOC 2GDPR · DPA99.5% uptime SLASpend limitsFree egressSSO
From
$0.15/GPU-hr
Billing
Per minute, balance updates in real time
Minimum commitment
None for on-demand
Egress
Free
Uptime SLA
99.5% monthly
Registered in
🇬🇧
Visit Hyperstack
On-demand pricing
×1, on demand · NVIDIA RTX Pro 6000 SE
$1.85/hr
Deploy
×1, on demand · NVIDIA H100 SXM
$3.20/hr
Deploy
×1, on demand · NVIDIA H200 SXM
$3.99/hr
Deploy
×1, on demand · NVIDIA A100 SXM
$1.60/hr
Deploy
×1, on demand · NVIDIA A6000
$0.50/hr
Deploy

See all 18 Hyperstack offers →

What’s good
Lowest published on-demand rate is $0.15/GPU-hr (NVIDIA RTX A4000 16GB), with no minimum commitment
Egress is free — no per-GB charge on data leaving
Billing is per minute
Self-serve signup — no sales call to get started
Publishes a 99.5% monthly uptime commitment, with service credits when it is missed
Free credits available: vouchers (single-use per organisation)
Plans
On-Demand VMsReservation contractsSpot VMs
From$0.15/GPU-hr (A4000)

Hyperstack · On-Demand VMs

Every on-demand rate we hold for Hyperstack, per GPU per hour, cheapest first.

×1, on-demand · NVIDIA A4000
$0.15/hr
Deploy
×1, on-demand · NVIDIA A6000
$0.50/hr
Deploy
×1, on-demand · NVIDIA L40
$1.00/hr
Deploy
×1, on-demand · NVIDIA A100 SXM
$1.60/hr
Deploy
×1, on-demand · NVIDIA RTX Pro 6000 SE
$1.85/hr
Deploy
×1, on-demand · NVIDIA H100 SXM
$3.20/hr
Deploy
×1, on-demand · NVIDIA H200 SXM
$3.99/hr
Deploy

Prices as last read from Hyperstack’s own pages. Open a card for every provider renting it.

From $0.11 (A4000)

Hyperstack · Reservation contracts

The rest of what Hyperstack lists for this plan, as its own page states it:

$1.75 (H100) · $2.72 (H100 SXM) · $2.79 (H200) · $5.10 (B200)

$0.40 (A6000)

Hyperstack · Spot VMs

Every spot rate we hold for Hyperstack, per GPU per hour, cheapest first.

×1, spot · NVIDIA A6000
$0.40/hr
Deploy
×1, spot · NVIDIA L40
$0.80/hr
Deploy
×1, spot · NVIDIA RTX Pro 6000 SE
$1.48/hr
Deploy
×1, spot · NVIDIA H100 PCIe
$2.00/hr
Deploy

Prices as last read from Hyperstack’s own pages. Open a card for every provider renting it.

Minimum termNoneDefined duration, agreed with salesNone
BillingPer minute, prepaid creditInvoiced monthlyPer minute
Uptime SLA99.5%99.5%No SLA
TenancyVMs, 1× to 8× (10× on some flavors)Reserved GPUs for a defined durationSurplus capacity, reclaimable without notice
ProvisioningMinutes, subject to live stockVia sales; guaranteed access to reserved stockSelf-serve when surplus exists
PaymentCredit card via Stripe (3D Secure); prepaid USDMonthly invoice; net 7 days unless agreedPrepaid credit
Details
Billing & access
How you pay
Prepaid by default; postpaid and premium postpaid via sales
Currency
USD; sales tax/VAT added where applicable
Commitment discounts
Reservation contracts with agreed discounts
Free credits
Vouchers (single-use per organisation)
Approval or KYC
Card plus KYC information required before use
When funds run out
Service stops; data kept 30 days then deletable; resources may be deleted after a week
Late payment
1.5%/month interest on overdue invoices; net 7 days
Provisioning guarantee
On-demand: none (shared pool); contracts guarantee reserved stock
What you get
GPUs listed
B300, B200, H200 SXM, H100 SXM, H100 NVLink, H100 PCIe, RTX Pro 6000 SE, A100 SXM, A100 NVLink, A100 PCIe, L40, RTX A6000, RTX A4000
Regions
NORWAY-1 (Vestland), CANADA-1 (Quebec), US-1 (Texas)
more
Norway and Canada are marked sustainably powered; additional regions may appear depending on account configuration.
Node configuration
Max pCPUs and RAM per GPU listed per model (e.g. H100 SXM 24 pCPU / 240 GB)
Interconnect
SR-IOV up to 350 Gbps inter-VM (CANADA-1, US-1); NVLink flavors
more
Standard networking in those regions is up to 125 Gbps; SR-IOV environments up to 350 Gbps for compatible flavors.
Datacenter tier
Tier 3
Local storage
Ephemeral and local NVMe per flavor
Persistent storage
Shared storage volumes $0.000096774/GB-hr (~$0.10/TB-hr); snapshots at Cloud-SSD rates
Ingress
Free
Public endpoints
Public IP $0.00672043/hr
Private networking
Environments per region; firewalls; bastion servers
Images
Ubuntu with NVIDIA drivers/CUDA (R535–R570), with or without Docker; AlmaLinux, Debian; no Windows
Access methods
Web console, API, SSH, VNC recovery console
Model deployment
AI Studio (inference, fine-tuning, playgrounds)
Managed inference
AI Studio — token-priced inference and fine-tuning
Managed Kubernetes
On-Demand Kubernetes, free master node
Kubernetes integration
On-Demand Kubernetes with CSI driver
Terraform
Terraform provider
FAQ
How much does Hyperstack cost?

As of September 2026, the lowest published on-demand rate is $0.15 per GPU-hour for the NVIDIA RTX A4000 16GB. Hyperstack publishes an on-demand rate for 7 GPU models. Billing is per minute.

Does Hyperstack have an uptime SLA?

Hyperstack states 99.5% monthly. Service credits: pro-rata account credit, claim within 5 days of cycle end, capped at fees for that service; terminate free after 3 consecutive missed months.

Is there a minimum commitment on Hyperstack?

No. Hyperstack states none for on-demand; prepaid balance.

Does Hyperstack charge for egress?

No. Hyperstack states that egress is free.

Can I sign up to Hyperstack without approval?

Hyperstack states that signup is self-serve. On approval and identity checks it states: card plus KYC information required before use.

Where are Hyperstack's data centres?

Hyperstack lists nORWAY-1 (Vestland), CANADA-1 (Quebec), US-1 (Texas).

Who do you contract with at Hyperstack?

The contracting entity is NexGen Cloud Limited (England and Wales).

Reviewed September 2026, from Hyperstack’s own site. 75 facts, each quoted from the page it was read on.