A Simple Guide to Europe's Complicated Cloud Provider Landscape

If you’re looking to run a large-scale compute project in the US, you’re spoiled for choice. There’s a wide range of capable cloud offerings, from IBM and Oracle to Cloudflare and DigitalOcean. But chances are, you’ll end up with one of the big three: Microsoft’s Azure, Google Cloud, or Amazon Web Services (AWS).

These platforms support massive scale, offer an incredible range of products and services, and provide straightforward setup and super-granular, usage-based billing.

The same is true for China, where Alibaba Cloud, Huawei, and Tencent have similarly rich and mature offerings.

But what if you’re in Europe? And - in a moment of unwavering patriotism or corporate necessity - decide to keep your products and services hosted by the home team?

Well, things are a lot less clear. But they’re not hopeless… and they’re getting better fast.

Where is Europe at, in general?

To some people’s surprise, Europe is not a country. And while the EU has a single market, Europe as a whole is far from being as integrated as the US or China. Different languages, regulations, procurement practices, and national interests still make it difficult for companies to expand across borders.

As a result, Europe hasn’t produced its own equivalent of the American cloud oligopoly. Instead, the continent has a much wider range of smaller, often nationally focused providers.

That’s not to say that Europe has escaped the dominance of AWS, Azure, and Google Cloud. Quite the opposite. What Europe lacks is a homegrown competitor operating at a comparable scale.

Still, there’s no shortage of alternatives. These can roughly be grouped into three categories.

General-Purpose Clouds

These are the AWS/Azure/GCP equivalents. At least, kind of…

They offer a broad range of products, from infrastructure-level services like compute, storage, and networking to platform-level offerings such as managed databases, Kubernetes clusters, message queues, serverless functions, and AI services.

Five of the most notable are:

• OVHcloud (France): One of Europe’s largest and most established cloud providers, with a broad portfolio of public and private cloud infrastructure, managed Kubernetes, databases, and storage.
• IONOS Cloud (Germany): A major German provider offering a comprehensive enterprise cloud platform with virtual data centers, managed databases, Kubernetes, and networking.
• Scaleway (France): A developer-oriented cloud with an increasingly impressive portfolio of managed services, serverless computing, databases, and AI infrastructure.
• STACKIT (Germany): The cloud platform of Schwarz Group, the company behind Lidl and Kaufland. Strongly focused on enterprise workloads, European sovereignty, and managed cloud services.
• T Cloud Public (Germany): Deutsche Telekom’s public cloud offering, formerly known as Open Telekom Cloud. A broad infrastructure and platform offering with a strong focus on enterprise and public-sector customers.

These providers are probably the closest Europe has to true hyperscaler alternatives. They offer many of the same fundamental building blocks, although their product portfolios, global reach, and ecosystems vary significantly.

Infrastructure Providers

Do you have your own DevOps guy and a good understanding of Kubernetes? Great! - because Europe has some offerings you’ll absolutely love for their pricing and simplicity.

Infrastructure providers focus on the fundamentals: virtual machines, dedicated servers, storage, networking, and sometimes managed Kubernetes or databases.

Rather than offering hundreds of tightly integrated cloud services, they give you the building blocks to run your own infrastructure. That means more operational responsibility, but often substantially lower costs and fewer surprises on your cloud bill.

Some of the most relevant players are:

• Hetzner (Germany): A favorite among developers for its exceptionally competitive pricing on virtual machines, dedicated servers, and storage.
• Leaseweb (Netherlands): A large international infrastructure provider offering bare-metal servers, public and private cloud, and extensive networking capabilities.
• UpCloud (Finland): A high-performance cloud infrastructure provider with virtual machines, storage, managed databases, and Kubernetes.
• Exoscale (Switzerland): Part of the Austrian A1 Group, offering cloud compute, storage, Kubernetes, and other developer-focused infrastructure services.
• Contabo (Germany): A budget-oriented infrastructure provider specializing in VPS, virtual dedicated servers, and bare-metal hosting.

These companies can be particularly attractive if you already know how to operate your own software stack and don’t need your cloud provider to manage every component for you.

Hosting Providers

If you’ve been hosting websites since the olden days, before the cloud emerged, you might remember renting a virtual host: essentially a folder on an Apache server, mapped to a domain name, with a small SQL database and PHP execution thrown in.

The companies offering those services are still around. And many have expanded into a much broader range of higher-level products, from managed WordPress and e-commerce hosting to website builders, email services, virtual servers, and cloud hosting.

Some of Europe’s most significant hosting companies include:

• Hostinger (Lithuania): A rapidly growing hosting provider focused on affordable website hosting, VPS, managed WordPress, and AI-powered website-building tools.
• team.blue (Belgium): A large pan-European hosting group operating dozens of local brands, including Combell, TransIP, and Register.it.
• Aruba (Italy): One of Italy’s most established hosting providers, offering domains, web hosting, dedicated servers, and, through Aruba Cloud, substantial enterprise cloud infrastructure.
• Infomaniak (Switzerland): A privacy-focused Swiss provider with web hosting, email, collaboration tools, and increasingly capable cloud infrastructure services.

The boundaries between these categories are admittedly blurry. IONOS and Aruba, for example, operate both traditional hosting businesses and general-purpose cloud platforms. Similarly, infrastructure providers like UpCloud and Exoscale increasingly offer managed platform services.

But the distinction is still useful: some providers sell you a complete toolbox, others give you the raw infrastructure to build your own, and others aim to make hosting applications as effortless as possible.

The Others

There’s also a growing range of specialized cloud providers focusing on AI, GPU computing, edge infrastructure, and other niches.

Companies like Nebius (Netherlands), Verda (Finland), Nscale (UK), and Mistral AI (France) are building increasingly ambitious AI infrastructure platforms.

But for this guide, we’ll stick to general-purpose clouds and infrastructure providers.

So, Where Are We At?

Compared with the big American and Chinese names, there are some notable differences in European cloud providers.

I’ll be generalizing a little, and you’ll probably find a counterexample for each of these claims. But I think they capture some important differences in how the European cloud market operates.

Technology

Do you love homegrown cloud services like Google Cloud Spanner, AWS Kinesis, or Microsoft Entra ID?

Well, you’re largely out of luck in Europe.

European cloud providers tend to rely much more heavily on established open-source technologies and commercial software, rather than building their own proprietary alternatives.

Want a managed database? You’ll probably get PostgreSQL, MySQL, or MongoDB. Need a message broker? Apache Kafka or RabbitMQ. Container orchestration? Kubernetes. Infrastructure management? Quite often OpenStack.

That’s not to say that European providers don’t build their own technology. Scaleway has developed its own serverless platform, IONOS offers a proprietary visual Data Center Designer, and virtually every serious cloud provider has its own control plane, provisioning systems, and management APIs.

But what Europe largely lacks is the enormous range of distinctive, deeply integrated, proprietary platform services offered by the American hyperscalers.

Now, this is a double-edged sword.

Bespoke services can provide entirely new functionality, higher-level APIs, and significant reductions in development time. But they also increase switching costs by tying you more closely to a particular cloud provider.

An application built around PostgreSQL, Kafka, and Kubernetes can, at least in principle, be moved between providers relatively easily. One deeply integrated with DynamoDB, Kinesis, and Lambda is going to have a much harder time leaving AWS.

Of course, there’s a catch: somebody still has to integrate and operate all that open-source software. And that somebody might well be you.

Flexibility

AWS lets you spin up a massive m8i.48xlarge EC2 instance, revel in the experience of having 192 virtual CPUs and 768 GiB of memory at your disposal, and shut the whole thing down half an hour later.

All for roughly $5.08 in its US East region, before storage and other charges.

That’s remarkable when you think about it.

And while European clouds have largely embraced pay-as-you-go pricing, the experience isn’t always quite as flexible.

Take Contabo, for instance. Its VPS products have a minimum contract duration of one month. That’s perfectly reasonable if you’re running a website around the clock, but less attractive if you just want a machine for a quick experiment.

Want to sign up for StackIT as a single developer? Oh dear, you need a company and proof that its located in Germany, Switzerland or Austria - and verification can take a while.

Other providers are considerably more flexible. Scaleway, for example, stops charging for compute when an instance is powered off, although attached storage and IP addresses continue to cost money. STACKIT offers hourly pay-as-you-go billing rather than requiring monthly contracts.

Then there’s the question of scale.

Hetzner’s largest general-purpose cloud instance offers 48 vCPUs and 192 GB of memory. Scaleway lists instance families reaching 64 vCPUs, while STACKIT goes considerably further, offering configurations with over 200 vCPUs.

So large instances certainly exist in Europe. But the available configurations, regional capacity, and ability to provision unusual amounts of compute on demand differ considerably between providers.

For typical workloads, you may never notice. But for large-scale analytics, temporary compute clusters, or applications with highly unpredictable demand, this flexibility can make an enormous difference.

Cost

Now here’s an area where European clouds can really shine.

Services like Hetzner, UpCloud, and OVHcloud offer infrastructure at prices that can make an AWS or Azure bill look downright extravagant.

Let’s take a concrete example.

An AWS c7i.xlarge instance with 4 vCPUs and 8 GiB of memory - the kind of workhorse you’d use for everyday servers - costs around $0.204 per hour in Frankfurt. Run it continuously for a 30-day month, and you’re looking at roughly $147, before storage and other charges.

A Hetzner CPX32, also offering 4 vCPUs and 8 GB of memory, costs just $41.99 per month in Germany.

That’s approximately 71% less, based purely on the advertised virtual-machine prices.

OVHcloud’s c3-8 instance, with 4 vCores and 8 GB of memory, costs approximately $0.108 per hour, or $78 over the same period - about 47% less than the AWS example, before additional storage and IP costs.

UpCloud also offers competitively priced small cloud servers, with its Starter range including a 4-vCPU, 8-GB configuration at €20 per month.

Now, these aren’t exact apples-to-apples comparisons. Hetzner’s CPX instances use shared CPU resources, while AWS’s c7i family is designed for sustained compute performance. Processor generations, storage, network performance, availability guarantees, and service levels can all differ.

And AWS offers substantial discounts through Savings Plans and Reserved Instances, which reduce the gap for predictable workloads.

Still, the basic point stands: if you’re running relatively straightforward infrastructure continuously, European providers can be considerably cheaper.

Another important difference is bandwidth pricing. Hetzner includes 20 TB of outbound traffic with many of its European cloud servers. AWS generally charges separately for internet data transfer beyond its free allowance.

If your application moves a lot of data, that difference can be just as important as the cost of the virtual machines themselves.

Of course, cheap servers aren’t the same thing as cheap software operations. If you need an additional DevOps engineer to maintain services that AWS would have managed for you, those infrastructure savings can disappear rather quickly.

Geographic Reach and Resilience

Here’s another area where size matters.

AWS operates across dozens of geographic regions worldwide, with multiple independent availability zones in each region. That makes it possible to build applications that survive individual data center failures, regional outages, and even substantial geographic disruptions.

European providers tend to operate on a much smaller geographic footprint.

STACKIT, for example, currently offers cloud regions in Germany and Austria, each with three availability zones. OVHcloud has a considerably larger international footprint, but many of its regions consist of only a single availability zone, while locations such as Paris and Milan offer three.

And that’s an important distinction.

Having three data centers isn’t necessarily the same as having three independent availability zones. And having multiple availability zones within one city isn’t the same as operating independent regions across different countries.

For a local business application, this might not matter much. For an international SaaS platform with demanding availability requirements, it certainly does.

It’s also worth checking whether the individual services you need are available in every region. A provider might offer virtual machines in ten locations but only provide its managed database or AI services in two.

Developer Ecosystem

One of the most overlooked advantages of AWS, Azure, and Google Cloud isn’t actually their technology.

It’s everything around it.

There are countless tutorials, libraries, Terraform modules, third-party integrations, monitoring solutions, deployment tools, and engineers who already know how to use their platforms.

Need to integrate some obscure service with AWS S3? Chances are, somebody has already written a library for it.

Need a Terraform example for deploying a fairly unusual infrastructure configuration? Somebody’s probably published one.

The European cloud ecosystem is smaller, although the reliance on established technologies helps considerably.

OpenStack APIs, S3-compatible object storage, Kubernetes, PostgreSQL, and standard Linux infrastructure all benefit from mature tooling that isn’t tied to any particular provider.

But once you venture beyond those fundamentals, you may find yourself working with provider-specific APIs, thinner documentation, fewer community examples, and integrations that haven’t quite caught up.

For experienced infrastructure engineers, this is often manageable. For smaller teams trying to assemble an entire production platform quickly, it can become a genuine disadvantage.

Sovereignty

And finally, there’s the reason you might have started looking at European clouds in the first place: sovereignty.

This is where European providers have an advantage that has very little to do with processor speed, API design, or pricing.

There is an important difference between keeping your data in Europe and keeping it under European control.

AWS, Azure, and Google Cloud all operate European data centers. You can host an application in Frankfurt, Paris, or Dublin without placing its primary data in an American data center.

But you’re still dealing with American companies, subject to American laws and potentially to legally binding demands from American authorities.

European-owned providers can reduce some of these jurisdictional dependencies. For governments, regulated industries, and businesses concerned about strategic independence, that’s an increasingly important consideration.

But European ownership doesn’t automatically guarantee complete technological sovereignty, either.

A provider might rely on American software, American hardware, non-European subcontractors, or technologies it doesn’t independently control.

The European Commission has started formalizing these aspects through its Cloud Sovereignty Framework, assessing providers across dimensions including legal control, operational independence, technological autonomy, and supply-chain dependencies.

And this points to something important.

For a long time, choosing a cloud provider was primarily an engineering and commercial decision. Increasingly, it’s also a geopolitical one.