GPUBank vs Hyperstack
Hyperstack is the GPU cloud of NexGen Cloud, a London company founded in 2020, that rents GPU virtual machines billed per minute from a prepaid balance in Canada, Norway and Texas. Its H100 SXM costs $3.20 per GPU-hour, sold only as an 8-GPU flavor; GPUBank charges $1.85 and rents a single GPU. Hyperstack has a wider catalogue, managed Kubernetes and SOC 2 Type II; GPUBank bills by the second, includes IPv4 and takes crypto.
GPUBank
Tallinn, Estonia
H100 SXM
$1.85 / h
- vCPU, RAM, NVMe, IPv4 and traffic included
- Per-second billing after the health check
- Top up in Bitcoin, USDT and 12+ other coins
vCPU, RAM, NVMe, IPv4 and traffic included
Hyperstack
London, United Kingdom · since 2020
H100 SXM
$3.20 / h
- 13 GPU models, including B300 on demand and RTX A4000 at $0.15 per GPU-hour
- Managed Kubernetes with a free master node
- SR-IOV networking up to 350 Gbps between VMs in CANADA-1 and US-1
Hyperstack
GPU cloud
United Kingdom27 points compared30 of their pages readprices last checked on 29 September 2026page updated 2 November 2026
The verdict
Wider catalogue and SOC 2 at Hyperstack. Lower prices and single H100 GPUs at GPUBank.
Hyperstack lists 13 GPU models, from the RTX A4000 at $0.15 per GPU-hour to the B300 at $7.40, with managed Kubernetes, SR-IOV networking up to 350 Gbps and a SOC 2 Type II attestation through its parent. On the six GPUs both of us rent, GPUBank costs 31 to 46 % less per GPU-hour. Hyperstack sells its H100 SXM, H200 and B200 as 8-GPU flavors only, bills a stopped VM in full unless it is hibernated, and charges for each public IPv4. It takes cards through Stripe only; GPUBank takes crypto and asks for no card.
HyperstackGPUBank 01The real cost
The same GPUs, side by side, per hour.
Every GPU model both of us rent, at each provider's own published price for one card, in US dollars, excluding VAT. Prices last checked on 29 September 2026.
| GPU | Hyperstack per hour | GPUBank per hour | Difference | 100 h at Hyperstack | 100 h at GPUBank |
|---|---|---|---|---|---|
 B200 SXM180 GB |
$6.00 | $4.15 | −31 % | $600.00 | $415.00 |
 H200 SXM141 GB |
$3.99 | $2.49 | −38 % | $399.00 | $249.00 |
 H100 SXM80 GB |
$3.20 | $1.85 | −42 % | $320.00 | $185.00 |
 H100 PCIe80 GB |
$2.50 | $1.35 | −46 % | $250.00 | $135.00 |
 RTX PRO 600096 GB |
$1.85 | $1.15 | −38 % | $185.00 | $115.00 |
 A100 SXM80 GB |
$1.60 | $0.95 | −41 % | $160.00 | $95.00 |
Hyperstack prices are per GPU; a flavor costs the rate times its GPU count. B200, H200, H100 SXM and A100 SXM come in 8-GPU flavors only. CPU, RAM and NVMe are included; IPv4 is extra.
A fine-tune, 40 h on an H100 SXM
Hyperstack
$128.00
GPUBank
$74.00
Saves $54.00 at GPUBank (−42 %)
A long-context job, 40 h on an H200
Hyperstack
$159.60
GPUBank
$99.60
Saves $60.00 at GPUBank (−38 %)
A training run, 100 h on an A100 SXM
Hyperstack
$160.00
GPUBank
$95.00
Saves $65.00 at GPUBank (−41 %)
02Their catalogue
Every GPU Hyperstack lists.
Hyperstack lists 13 GPU models. The 11 below are the ones recorded in our check, each at its lowest public price per GPU-hour. 6 of them are also rented at GPUBank (in blue).
- B300288 GB · on demand$7.40
- B200192 GB · on demand$6.00
- H200141 GB · on demand$3.99
- H100 SXM80 GB · on demand$3.20
- H100 PCIe80 GB · on demand$2.50
- A100 SXM 80GB80 GB · on demand$1.60
A100 PCIe 80GB80 GB · on demand$1.35
L4048 GB · on demand$1.00- RTX PRO 600096 GB · on demand$1.85
RTX A600048 GB · on demand$0.50
RTX A400016 GB · on demand$0.15
03What the price includes
What appears on the bill besides the GPU.
The same H100 SXM session of 2 h 14 m 03 s, line by line, as each provider describes its billing.
Hyperstack
2 h 14 m 03 s on H100 SXM
- GPU · $3.20 / h
- $7.15
- vCPU and RAM
- Included
- Storage
- $0.000097 per GB-hour, about $0.07/GB-month
- Egress
- Free
- Public IPv4
- $0.0067 per hour, about $5 per month
- Billing granularity
- Per minute
- Minimum charge
- None per launch
GPUBank
2 h 14 m 03 s on H100 SXM
- GPU · $1.85 / h
- $4.13
- vCPU and RAM
- Included
- Storage
- Local NVMe included; persistent disk $0.03 per GB-month
- Egress
- Included
- Public IPv4
- Included
- Billing granularity
- Per second, from the health check
- Minimum charge
- None
While stopped
What keeps running on the bill.
Hyperstack
Stopped VM billed in full; hibernation bills disk.
A hibernated VM pays about $0.07 per GB-month for its root disk. The ephemeral disk is wiped.
GPUBank
Only the persistent disk, $0.03 per GB-month. Compute, IPv4 and scratch cost nothing.
04Where the GPUs run
3 countries against 4, and where they are.
Countries where Hyperstack says its GPUs run, from its own region list or documentation, against GPUBank's four regions. A pin marks a country, not the exact site.
Canada
Norway
United States
Paris
Frankfurt
Ashburn
Singapore
Hyperstack3
Canada
Norway
United States
GPUBank4
Paris
Frankfurt- Ashburn
Singapore
Hyperstack · 3 countriesGPUBank · 4 regions
05Paying and signing up
How each one takes your money.
Hyperstack
Payments and sign-up
Payment methods
Cards
Bank transfer
No cryptoMinimum top-up Not published
To start
- 1Email and password, or SSO
- 2Billing information to activate the account
- 3Card via Stripe with 3D Secure
- 4Age 18+
Billing modelPrepaid; postpaid for approved accounts
GPUBank
Payments and sign-up
Payment methods
A prepaid USD balance, topped up in crypto. Every GPU-second is drawn from it.
Minimum top-up $25
To start
- 1Email and password
- 2Top up in crypto
- 3Launch a GPU
Billing modelPrepaid balance
06Trust and support
Certificates, uptime and who answers.
| Hyperstack | GPUBank | |
|---|---|---|
| SOC 2 | Type II | Not published |
| ISO 27001 | In progress | Not published |
| HIPAA | Planned | No |
| GDPR DPA | DPA available | Published DPA, applies automatically |
| Availability target | 99.5 % monthly, credits on claim | 99.9 % monthly, per region |
| Status page | status.hyperstack.cloud | Public status page |
| Support | Email, tickets and live chat | Tickets from a $25 balance |
07Their site, our site
What you see before you pay.
Both sites as a visitor sees them, without an account: the Hyperstack pricing page captured on 29 September 2026, and GPUBank's live board. Cookie banners and chat bubbles are closed; nothing else is edited.
- 1The pricing page shows per-GPU hourly prices without an account, in on-demand, reservation and spot tables.
- 2Reservation prices are marked Starting from, with no term or minimum quantity stated.
- 3The payment method appears only in an expanded FAQ, which says payments go through Stripe by credit card.
- 1Live stock per region and every price, without an account.
- 2Per-hour and per-second prices, with vCPU, RAM, NVMe, IPv4 and traffic in them.
- 3A machine-readable price feed at /prices.json.
08The spec sheet
27 points, read off both sites.
GPUBank aheadHyperstack aheadEven
| Criterion | GPUBank | Hyperstack |
|---|---|---|
| Price and catalogue | ||
| H100 SXM, per GPU-hour | $1.85, all-in | $3.20, 8-GPU flavor only |
| Cheapest GPU-hour | $0.23, RTX 4090 | $0.15, RTX A4000Ahead |
| GPU models listed | 10 | 13Ahead |
| What the GPU price includesBoth include compute and local disk in the GPU price. Hyperstack bills IPv4 separately. | vCPU, RAM, NVMe, IPv4, traffic | CPU, RAM, root and ephemeral NVMe |
| Billing | ||
| Billing granularity | Per second | Per minute |
| Minimum chargeTo deploy, the balance must cover all running resources plus the new VM for one hour. | None | None per launch |
| Stopped instanceA hibernated VM pays about $0.07 per GB-month for its root disk. The ephemeral disk is wiped. | Disk only | Stopped VM billed in full; hibernation bills disk |
| Persistent storage | $0.03 per GB-month | $0.000097 per GB-hour, about $0.07/GB-month |
| Egress traffic | Included | Free |
| Public IPv4 | Included | $0.0067 per hour, about $5 per month |
| Billing modelMonthly invoices for reservations and sales-approved accounts, due in 7 days. | Prepaid balance | Prepaid; postpaid for approved accountsAhead |
| Payment and sign-up | ||
| Crypto payments | Bitcoin, USDT, Ethereum and 12+ more | Not accepted |
| Card payments | No | Cards via Stripe with 3D SecureAhead |
| Minimum deposit | $25 | Not published |
| Platform | ||
| API and CLI | REST API, CLI, Python SDK | REST API and alpha SDKs; no CLI documented |
| Terraform | Official provider | Official provider, labelled alpha |
| Serverless GPUsAI Studio sells token-priced inference and fine-tuning, not serverless GPU containers. | No | No |
| Managed Kubernetes | No | Managed, free master nodeAhead |
| Ready-made templates | 25 | Not published |
| Interruptible or spotGPUBank interruptible instances are 40 % below on demand. | Yes, 40 % off | Spot VMs at 20 % below on demand, 5 models |
| GPUs per instance | Up to 8 | Up to 10 per VM; clusters on requestAhead |
| Multi-node interconnect | Not offered | NVLink; SR-IOV up to 350 Gbps between VMsAhead |
| Network and regions | ||
| Regions | 4 | 3 |
| Countries | France, Germany, US, Singapore | Canada, Norway, United States |
| EU regions | Paris, Frankfurt | None in the EU; Norway serves RTX A4000 only |
| Trust and support | ||
| Availability targetGPUBank's terms set a 99.9 % monthly target per region. | 99.9 % monthly | 99.5 % monthly, credits on claim |
| Compliance reports | None published | SOC 2 Type II via NexGen CloudAhead |
| Public status page | Yes | status.hyperstack.cloud |
| Support | Tickets from a $25 balance | Email, tickets and live chatAhead |
09The right fit
Different needs. A clear choice.
What Hyperstack does better
- 13 GPU models, including B300 on demand and RTX A4000 at $0.15 per GPU-hour
- Managed Kubernetes with a free master node
- SR-IOV networking up to 350 Gbps between VMs in CANADA-1 and US-1
- SOC 2 Type II through NexGen Cloud and a 99.5 % SLA with credits
- Monthly invoicing for reservations and approved postpaid accounts
Choose GPUBank if
- H100 SXM at $1.85 per GPU-hour, against $3.20 on an 8-GPU flavor
- Rent one H100 SXM or H200; Hyperstack sells them in 8-GPU flavors only
- Per-second billing instead of per minute
- Public IPv4 included; Hyperstack charges $0.0067 per hour
- Pay in Bitcoin, USDT and 12+ other coins; regions in Paris and Frankfurt
Moving from Hyperstack takes an afternoon.
Hyperstack charges nothing for egress, so copying data out is free on its side. Its ephemeral disks are wiped on hibernation or deletion, so copy data before leaving.
- 1Launch the same GPU on GPUBank with a persistent disk in the region you need.
- 2Copy volumes over SSH with rsync, or pull object storage with standard S3 tools.
- 3Run your Docker image or a matching template, then delete Hyperstack VMs and volumes.
# 1. a GPU and a disk on GPUBank gpubank launch --gpu h100-sxm --region eu-paris --disk data # 2. copy your data from the old machine rsync -azP old-host:/workspace/ root@<new-ip>:/workspace/ # 3. run the same image or template ssh root@<new-ip> 'cd /workspace && ./start.sh'
A few things you might be wondering.
Yes. Hyperstack lists the H100 SXM at $3.20 per GPU-hour, sold only as an 8-GPU flavor. GPUBank charges $1.85 per GPU-hour and rents a single GPU, with vCPU, RAM, NVMe, IPv4 and traffic included.
Yes. A stopped (SHUTOFF) VM is billed in full. Only hibernation stops GPU billing, and it keeps the root disk at about $0.03 per GB-month while wiping the ephemeral disk. On GPUBank a stopped instance costs only its persistent disk, at $0.03 per GB-month.
No. Self-serve payments go through Stripe by card with 3D Secure. GPUBank is paid from a USD balance topped up in Bitcoin, USDT, Ethereum and more than a dozen other coins.
Hyperstack runs NORWAY-1 in Vestland, Norway, which serves RTX A4000 flavors only. Its other regions are in Quebec and Texas. GPUBank has regions in Paris and Frankfurt, plus Ashburn and Singapore.
Hyperstack. It offers SR-IOV networking up to 350 Gbps between VMs in CANADA-1 and US-1, managed Kubernetes and dedicated clusters on request. GPUBank rents up to 8 GPUs per instance and has no managed Kubernetes.
Without. Every price on this page excludes VAT and sales tax, for Hyperstack and for GPUBank, so the two columns compare like for like. Tax is added at checkout according to where the buyer is based.
Sources
Every figure about Hyperstack was read on its own pages or official documentation. Prices last checked on 29 September 2026.
- 01hyperstack.cloud/gpu-pricingon-demand, reservation and spot GPU prices, storage, IP, egress, CPU flavors, AI Studio prices, FAQ (Stripe payments, per-minute billing, balance drain)
- 02hyperstack.cloud/site observations, regions claim, product list
- 03docs.hyperstack.cloud/docs/hardware/flavorsflavor catalogue: GPU counts, CPU, RAM, disks, regions, spot flavors, networking
- 04docs.hyperstack.cloud/docs/billing/pricebookper-GPU billing model, included CPU/RAM/disk, hibernation billing, Pricebook API
- 05docs.hyperstack.cloud/docs/billing/billing-policiessign-up methods, prepaid/postpaid, per-minute billing, Stripe 3DS, minimum balance rule, low-balance actions
- 06docs.hyperstack.cloud/docs/billing/states-and-billingstopped vs hibernated billing, root disk and IP rates
- 07docs.hyperstack.cloud/docs/billing/snapshotssnapshot storage billing
- 08docs.hyperstack.cloud/docs/billing/contractsreservation contracts
- 09docs.hyperstack.cloud/docs/resource-management/regionsregions, locations, Tier 3, SOC 2 Type 2, SR-IOV availability
- 10docs.hyperstack.cloud/docs/virtual-machines/spotspot terms and limitations
- 11hyperstack.cloud/spot-vmsconflicting spot price table (site observation only)
- 12docs.hyperstack.cloud/docs/storage/object-storageobject storage price and region
- 13docs.hyperstack.cloud/docs/network/public-ippublic IP billing and behaviour
- 14docs.hyperstack.cloud/docs/network/high-speed-networkinginterconnect
- 15docs.hyperstack.cloud/docs/hardware/gpu-stock-informationlive stock on 2026-09-29, spot flavors in stock
- 16docs.hyperstack.cloud/docs/hardware/gpu-cluster-requestclusters
- 17docs.hyperstack.cloud/docs/libraries/SDKs and Terraform (alpha)
- 18docs.hyperstack.cloud/docs/libraries/terraformTerraform
What we could not verify (8)
- Spot prices for H100 SXM, H100 NVLink and A100 SXM: spot flavors exist (flavors/stock pages) but the pricing page spot table does not list them; the Spot VMs page figures conflict with current on-demand prices, so they were not recorded
- Reservation term lengths and minimum quantities behind 'Starting from' prices not published
- B300 reservation price and GB200/GB300 NVL72 prices not published (sales only)
- Minimum top-up amount not published
- No official CLI documented (API, SDKs, Terraform, MCP server only)
- Exact data-center cities not published (only Quebec, Vestland and Texas)
- Funding and ownership details of NexGen Cloud not read on an official page
- Support gating (minimum spend for tickets/chat) not published
In short
GPUBank vs Hyperstack, in facts.
The essentials of this comparison in one table, for quick readers and AI assistants. Prices last checked on 29 September 2026.
Read as Markdown- H100 SXM, per GPU-hour
- Hyperstack: $3.20, 8-GPU flavor only
GPUBank: $1.85, all-in - Cheapest GPU-hour
- Hyperstack: $0.15, RTX A4000
GPUBank: $0.23, RTX 4090 - Billing granularity
- Hyperstack: Per minute
GPUBank: Per second - Stopped instance
- Hyperstack: Stopped VM billed in full; hibernation bills disk
GPUBank: Disk only - Persistent storage
- Hyperstack: $0.000097 per GB-hour, about $0.07/GB-month
GPUBank: $0.03 per GB-month - Egress traffic
- Hyperstack: Free
GPUBank: Included - Crypto payments
- Hyperstack: Not accepted
GPUBank: Bitcoin, USDT, Ethereum and 12+ more - Card payments
- Hyperstack: Cards via Stripe with 3D Secure
GPUBank: No - Minimum deposit
- Hyperstack: Not published
GPUBank: $25 - GPU models listed
- Hyperstack: 13
GPUBank: 10 - Regions
- Hyperstack: 3
GPUBank: 4 - Serverless GPUs
- Hyperstack: No
GPUBank: No - Availability target
- Hyperstack: 99.5 % monthly, credits on claim
GPUBank: 99.9 % monthly - Compliance reports
- Hyperstack: SOC 2 Type II via NexGen Cloud
GPUBank: None published - Support
- Hyperstack: Email, tickets and live chat
GPUBank: Tickets from a $25 balance - Headquarters
- Hyperstack: London, United Kingdom
GPUBank: Tallinn, Estonia
How this page was made
Every price and policy of Hyperstack was read on its own website, documentation or public API on 29 September 2026, with the address of each page listed above. A value we could not read on a primary source is marked "Not published" rather than estimated. GPUBank figures come from this site and its terms. Prices change: check both providers before you buy. Hyperstack, its logo and its website belong to their owner and are shown only to identify the service compared. GPUBank is not affiliated with Hyperstack.
