GPU Prices in Europe: What an H100 Hour Costs

NVIDIA H100 pricing in Europe runs from €2.80 ($3.13) to €7.98 ($8.93) per GPU hour for a single GPU, depending on which GPU cloud you rent it from. OVHcloud and Scaleway are under €3, while AWS and Azure charge more than double in their EU regions. Rented by the hour, German providers start at €5.22 ($5.84) for an H100 (T Cloud Public); AIME's monthly plan works out to about €3.88 ($4.33) per hour. IONOS rents out the H200 in Germany for €3.00 ($3.36) per hour, the cheapest H200 in this comparison.

Every price in this article comes straight from the providers' own price APIs and price lists. AWS and Nebius only publish US dollar prices, which we convert to euros at the ECB reference rate. We refresh the tables every month.

We compare the on-demand list price per GPU hour, excluding VAT: one hour with no contract. Reservations and commitments are structured differently by provider and term, so they are left out. Disclosure: Kontinent itself offers dedicated inference on GPUs in the EU.

Key takeaways

• H100 pricing from €2.80 ($3.13) per hour: OVHcloud and Scaleway stay under €3; Verda is at €3.48 ($3.89) after its price increase.

• At German providers: the H200 at IONOS for €3.00 ($3.36), the H100 at T Cloud Public for €5.22 ($5.84) and at STACKIT for €8.04 ($9.00) per GPU hour in an eight-GPU server. AIME in Berlin rents an H100 NVL for €2,829 ($3,165) a month.

• Hyperscalers charge more than double: Azure asks €7.98 ($8.93) for an H100 NVL in Frankfurt, AWS €6.58 ($7.36) in Stockholm.

• AWS lists no H100 in Frankfurt: eu-central-1 carries no P-series EC2 instance newer than the A100. The nearest AWS H100 in the EU is in Stockholm.

• Azure has no 8-GPU H100 nodes in Germany: in the EU, the ND H100 v5 series is only listed in the Netherlands, Poland, Sweden and Spain.

• Nebius and Verda got more expensive: since October 1 the H100 costs $4.50 instead of $3.85 at Nebius (+17%). At Verda it rose from €2.95 to €3.48 between September 24, 2026 and October 8, 2026 (+18%).

• Managed costs extra: Scaleway's Managed Inference on the same H100 costs €3.40 ($3.80) instead of €2.87 ($3.21), about 19% more.

Table of contents

1. H100 pricing per GPU hour

2. Renting a GPU server in Germany

3. H200, B200 and B300 pricing

4. A100, L40S and L4 for smaller models

5. Where the GPU runs: regions and headquarters

6. Nebius and Verda raised their prices

7. What a GPU server costs per month

8. Rent, reserve or run dedicated GPU servers

9. Methodology

10. Work out your GPU setup with us

11. Frequently asked questions

12. Sources

H100 pricing per GPU hour

Every provider in this table rents out the H100 as a cloud instance, meaning a GPU server billed by the hour. It comes as an 80 GB PCIe card, as an SXM module with NVLink across every GPU in a node, and as the H100 NVL with 94 GB, which is also a PCIe card. NVIDIA lists 3.35 TB/s of memory bandwidth at up to 700 W for the SXM H100, 3.9 TB/s at 350 to 400 W for the H100 NVL and 2 TB/s at 350 W for the 80 GB H100 PCIe. So the NVL has more memory bandwidth than SXM, but less FP8 compute (3,341 vs 3,958 TFLOPS with sparsity), and NVLink only bridges it in pairs. The cheapest SXM H100 in this comparison is Scaleway at €3.17 ($3.54) per GPU hour in a node with 8 GPUs.

If you want an SXM H100 at Azure, you have to take a node with eight GPUs and pay for all eight at €14.06 ($15.73) each. At Scaleway, the eight-GPU SXM node is the cheapest per GPU; smaller instances cost slightly more.

Renting a GPU server in Germany

The table above covers the large European GPU clouds and the hyperscalers. If you want to rent the GPU server from a German company, there is a second group: IONOS, Deutsche Telekom's T Cloud Public (formerly Open Telekom Cloud), STACKIT from the Schwarz Group, AIME in Berlin, plus Hetzner and centron for smaller GPUs. None of them publishes prices in a format our script reads automatically, so we read them from each provider's pricing page or price API on October 8, 2026.

Renting a GPU server by the hour: H100 and H200

At €3.00 ($3.36) per hour, IONOS has the cheapest H200 in this article, cheaper even than any hyperscaler H100. It is the PCIe variant: less compute than the SXM H200 and no NVLink across the whole node, which rarely matters for inference on one or a few GPUs. STACKIT documents its H100 as an eight-GPU server at €64.33 ($71.96) per hour. Its price API also lists smaller H100 servers at the same price per GPU, which the docs do not describe. T Cloud Public and STACKIT only give a region code in their price sources, not a city.

H100 and H200 by the month

AIME does not bill by the hour. The shortest term is one week, and a six-month term lowers the monthly price. Spread over 730 hours, its monthly H100 NVL works out to about €3.88 ($4.33) per hour, below Azure in Frankfurt.

Dedicated GPU servers at Hetzner, IONOS, centron and others

For inference on smaller models, image generation or testing, a workstation GPU is often enough. Hetzner lists both an hourly and a monthly price for its GPU servers and charges a one-time setup fee. According to the product pages, the GEX45 and GEX63 can currently only be ordered in Helsinki. Where a provider only gives an hourly price, we multiply it by 730 hours for the month.

IONOS lists its dedicated GPU servers including VAT; we convert them to net at 19%. On the same pricing page, centron says once that vCPU, RAM and storage come on top of the GPU price and once that they are included. We show the price as published.

H200, B200 and B300 pricing

The H200 has 141 GB of HBM3e instead of 80 GB. That matters for large models and long contexts, where VRAM runs out before compute does.

H200

Azure charges €12.13 ($13.56) per GPU hour for the H200 in Frankfurt, but only in a node with eight SXM GPUs. The cheapest single H200 in this table is Verda at €4.54 ($5.07). IONOS rents the H200 as a PCIe card in Germany for €3.00 ($3.36), about a quarter of the Azure price.

B200 and B300

B200 and B300 are the Blackwell generation; NVIDIA calls the B300 Blackwell Ultra. VRAM differs by offer: 180 GB for the B200, 268 to 288 GB for the B300 depending on the provider. According to its own region overview, Nebius offers neither in a public EU region: its B200 runs in Israel and the US, its B300 in the UK.

A100, L40S and L4 for smaller models

Not every project needs an H100. For models up to about 8B parameters and for LoRA fine-tuning, smaller cards are often enough.

The strongest P-series training GPU AWS lists in Frankfurt is the A100, as p4de with eight GPUs at €3.83 ($4.29) per GPU hour.

Where the GPU runs: regions and headquarters

If your data has to stay in Germany, or at least in the EU, the region matters as much as the price. Two findings from the price lists affect the architecture directly:

• AWS Frankfurt (eu-central-1): the price list carries no P-series EC2 instance newer than the A100 (p4d, p4de) there, so no H100, H200 or B200. Of the EU regions we query, AWS has the H100 only in Stockholm and the H200 in Stockholm and Spain. Outside the EU, London also has H100 instances.

• Azure Germany West Central: the H100 is only available there as the NC series with one or two H100 NVL. In the EU, the ND H100 v5 series with eight SXM GPUs for distributed training is only listed in the Netherlands, Poland, Sweden and Spain. Azure does offer ND nodes with eight H200s in Frankfurt.

The headquarters column shows where the parent company is based. AWS and Microsoft are US companies and fall directly under the US CLOUD Act (18 U.S.C. § 2713), even for data in EU data centers. For European providers with US subsidiaries or US regions, it depends on the corporate structure. Our take on sovereign AI in Europe is on our about page. More in Training AI models under GDPR and the AI Act.

Nebius and Verda raised their prices

Nebius and Verda have raised their list prices noticeably since our previous retrieval on September 24, 2026, Verda for every GPU in this comparison. Azure's euro prices are about 2% higher, with the H100 NVL in Frankfurt at €7.98 instead of €7.79. Nebius had announced the new prices on its pricing page for October 1, 2026. For Verda, the comparison of our two retrievals shows the increase. Nebius is shown in US dollars and Verda in its own euro price list, so the exchange rate does not distort the change.

That puts the H100 at €4.02 ($4.50) per GPU hour at Nebius and €3.48 ($3.89) at Verda. Both stay well below AWS and Azure, but OVHcloud and Scaleway are now clearly cheaper.

What a GPU server costs per month

A GPU running around the clock adds up to roughly 730 hours a month. At list price, a single H100 costs:

For distributed training, what matters are nodes with eight SXM GPUs connected by NVLink:

For a single GPU, the difference between the cheapest EU provider and Azure in Frankfurt is a little over €3,700 ($4,200) a month, and a little over €3,300 ($3,700) against Azure in Ireland. For an eight-GPU SXM node, it is about €64,000 ($71,000) between Scaleway and Microsoft Azure.

Rent, reserve or run dedicated GPU servers

List prices carry no discount at all. AWS and Azure discount through Savings Plans and reservations, and Nebius states up to 35% off for multi-month reservations of large clusters on its pricing page. Spot instances at Verda cost half the on-demand price but can be reclaimed at any time. That suits batch jobs, less so production inference.

Managed offerings fold operations into the hourly price. At Scaleway, the same H100 costs €3.40 ($3.80) per hour as Managed Inference and €2.87 ($3.21) as a bare instance.

A model that serves requests all day needs reserved GPUs configured for the workload, and someone to run them. That is what we offer as dedicated inference. If you want to train first, LLM fine-tuning in Europe lays out the four routes, and post-training on EU infrastructure shows how we run it for you.

Methodology

• As of: prices retrieved on October 8, 2026.

• Sources: OVHcloud order catalog API, Scaleway's GPU pricing page data (exact per-minute price times 60), Nebius pricing page, region overview and VM preset list, Verda instance-types API (api.datacrunch.io), AWS EC2 Price List, Azure Retail Prices API.

• German providers: read on October 8, 2026 from IONOS (Cloud GPU VM and GPU Server pricing pages), T Cloud Public (pricing calculator price API, API cache of September 17, 2026), STACKIT (SKU price API and machine type docs), AIME (GPU cloud pricing page), Hetzner (GPU server price data, without the €1.70/month IPv4 address) and centron (cloud GPU pricing page, as of July 16, 2026).

• Regions queried: AWS eu-central-1, eu-north-1, eu-south-1, eu-south-2, eu-west-1 and eu-west-3 in the EU, plus eu-west-2 (London) outside the EU; Azure every EU region where the Retail Prices API lists the VM.

• Conversion: AWS and Nebius only publish US dollar list prices; we show those as published and convert them to euros at the ECB reference rate of October 8, 2026, €1 = $1.1186. Verda publishes its own euro and US dollar prices, and both are shown as published. Microsoft Azure, OVHcloud and Scaleway publish euro prices; their US dollar figures in brackets are converted at the same ECB rate.

• Calculation: the instance's on-demand list price divided by its number of GPUs. Per provider and GPU variant we show the cheapest EU instance, plus Frankfurt for Azure.

• Not included: VAT, storage, data transfer, reservation and volume discounts, spot prices.

• Not compared: Google Cloud, because Google only publishes machine-readable prices through the Cloud Billing Catalog API with an API key. Genesis Cloud, because its website was unreachable on the day of retrieval.

• Listed does not mean available: an instance in a price list is not automatically bookable right away.

• ¹ OVHcloud lists GPU instances in Gravelines, Limburg, London and Warsaw on its availability page, but does not say which model runs where.

• ² Verda (formerly DataCrunch, based in Helsinki) names EU locations but not a location per GPU.

• ³ VAT not stated: T Cloud Public, STACKIT, AIME and Hetzner do not say in these sources whether prices are net or gross. We show the prices as published. IONOS Cloud publishes net prices, IONOS dedicated GPU servers gross (converted to net by us). USD figures for German providers are converted at the ECB rate, except Hetzner, which publishes its own USD list.

Work out your GPU setup with us

Which GPU fits depends mostly on utilization and where your data is allowed to live. We work it out with you in concrete numbers. Book a demo

Frequently asked questions

How much does an H100 cost per hour in Europe?

A single H100 costs between €2.80 ($3.13) at OVHcloud and €7.98 ($8.93) at Azure in Frankfurt, as the on-demand list price per GPU hour excluding VAT, as of October 8, 2026. In an Azure node with eight SXM GPUs it is €14.06 ($15.73) per GPU.

Where can I rent a GPU server in Germany?

By the hour with a German location: the H200 at IONOS for €3.00 ($3.36) per GPU hour and the H100 NVL at Azure in Frankfurt for €7.98 ($8.93). German providers with the H100 also include T Cloud Public (€5.22 ($5.84), region eu-de) and STACKIT (€8.04 ($9.00) in an eight-GPU server, region eu01). By the month, AIME in Berlin rents an H100 NVL for €2,829 ($3,165). AWS lists no H100 as an EC2 instance in Frankfurt.

How much does a GPU server cost per month?

Running around the clock (730 hours), an H100 at list price costs from about €2,044 ($2,286) a month at OVHcloud up to €5,829 ($6,520) at Azure in Frankfurt. Smaller GPUs cost far less: Hetzner rents a dedicated server with an RTX PRO 4000 for €212 ($247) a month plus a €209 ($249) setup fee.

Can I rent a GPU server by the hour?

Yes. Every cloud provider in the H100 and H200 tables except AIME publishes hourly prices, IONOS Cloud bills by the minute according to its pricing page, and centron by the hour with no setup fee. Hetzner lists an hourly price but charges a one-time setup fee. At AIME the shortest term is one week.

How much does a GPU server cost at Hetzner?

The GEX45 with an RTX PRO 4000 costs €0.34 ($0.40) per hour or €212 ($247) a month plus a €209 ($249) setup fee, and according to the product page can currently be ordered in Helsinki. The GEX131 with an RTX PRO 6000 costs €2.40 ($2.72) per hour or €1,497 ($1,697) a month and can also be ordered in Falkenstein according to its product page. Hetzner does not say on these pages whether prices are net. As of October 8, 2026.

How much does a GPU server cost at IONOS?

IONOS Cloud rents the H200 PCIe in Germany for €3.00 ($3.36) per GPU hour net, billed by the minute. The dedicated GPU Server IP4-50 with an RTX PRO 6000 works out to €2.10 ($2.35) per hour net, at most €1,513 ($1,692) a month, plus a setup fee of the same amount.

Why is the H100 so much more expensive at AWS and Azure?

Hyperscaler instances include more, such as large amounts of RAM, local NVMe and fast networking between nodes. Large customers also pay less than list price through Savings Plans and enterprise agreements.

Is the H100 NVL the same as an H100 SXM?

No. The H100 NVL is a PCIe card with 94 GB and, according to NVIDIA, 3.9 TB/s of memory bandwidth, more than the SXM variant with 80 GB and 3.35 TB/s. The SXM H100 has more FP8 compute and NVLink across every GPU in the node, while NVLink only bridges the NVL in pairs. For distributed training, SXM is the better choice.

Do the prices include VAT?

No. Prices exclude VAT where the provider states it, and exclude storage and data transfer. T Cloud Public, STACKIT, AIME and Hetzner do not say whether their prices are net or gross; they are marked in the tables.

Is buying a GPU cheaper than renting?

It depends on utilization. At 730 hours a month, a rented H100 at €2.80 ($3.13) costs a little over €24,500 ($27,400) a year. The full calculation with purchase, power and operations is in LLM fine-tuning in Europe.

How often are the prices updated?

Monthly. The tables are generated by a script straight from the providers' price sources; the retrieval date and exchange rate are listed under Methodology.

Why is Google Cloud missing?

Google only publishes machine-readable prices through the Cloud Billing Catalog API with an API key. We will add Google Cloud once we check its prices automatically.

Sources

• OVHcloud Public Cloud Order Catalog

• OVHcloud: Regions availability

• Scaleway: GPU Instances pricing

• Nebius: AI Cloud pricing

• Nebius: Regions

• Verda: Pricing

• AWS: EC2 On-Demand Pricing

• AWS: EC2 instance types by Region

• Microsoft: Azure Retail Prices API

• NVIDIA: H100 Tensor Core GPU

• NVIDIA: H100 PCIe GPU Product Brief

• Nebius: VM types and presets

• Verda: Instance types API (EUR)

• IONOS: Cloud GPU VM

• IONOS: GPU Server

• T Cloud Public: Price API

• STACKIT: SKU price API

• STACKIT: Machine types

• AIME: GPU Cloud

• Hetzner: GPU servers

• centron: Cloud GPU

• Cornell LII: 18 U.S.C. § 2713 (CLOUD Act)

• ECB: Euro foreign exchange reference rates

As of October 8, 2026 · prices from the providers' own price sources, ECB rate of October 8, 2026 · kontinent.ai · Not legal advice.