Hetzner Cloud vs Dedicated: A Guide to Saving Money

Your cloud hosting bill is probably higher than it needs to be. We often default to cloud servers because they are flexible and easy to start. But that flexibility comes at a cost a literal one that you pay every month.

Hetzner offers a powerful alternative that many overlook their dedicated auction servers. Understanding when to use these instead of cloud servers is one of the simplest ways to significantly reduce your infrastructure costs without sacrificing performance. In many cases you actually gain performance.

What Are Hetzner Auction Servers?

The Hetzner Server Auction or Serverbörse is a marketplace for their used dedicated servers. When a server is no longer needed by a customer it gets tested thoroughly by Hetzner and then placed in the auction. These are not weak or old machines. You can often find powerful multi core servers with lots of RAM and fast NVMe storage.

The key features are a low fixed monthly price and no setup fee. You are renting a physical machine. You get all the CPU all the RAM and all the disk I/O for yourself. This is fundamentally different from a cloud server which is a virtual machine running on shared hardware.

The Case for Dedicated Servers

The most obvious reason to choose a dedicated server is predictable cost. You pay one flat fee per month. There are no surprise charges for CPU usage or network traffic. This makes budgeting simple and stress free. If you have a service that runs 24/7 a fixed monthly cost is almost always cheaper than paying by the hour.

The second reason is raw power. On a cloud server you are sharing the underlying CPU with other customers. A 'vCPU' is not a real CPU core. On a dedicated server you get exclusive access to all the physical cores. For workloads that are CPU intensive like compiling code running CI/CD jobs or processing data a dedicated server will be dramatically faster than a similarly priced cloud instance.

Databases are a perfect example. A PostgreSQL or MySQL database loves fast disk I/O and direct CPU access. Running your database on a dedicated server with NVMe drives can give you a huge performance boost over a cloud server where I/O can be inconsistent.

The Case for Cloud Servers

Cloud servers still have their place. Their primary advantage is flexibility. You can create or destroy a cloud server in seconds using an API. This is something you cannot do with a dedicated server which can take an hour or more to provision.

This speed makes cloud servers ideal for workloads with unpredictable or spiky traffic. Imagine your website gets featured on a popular blog. You can use an API to quickly spin up ten more web servers to handle the load and then shut them down when the traffic subsides. You only pay for the few hours they were running.

Cloud servers also come with useful features that are managed for you. Things like snapshots for quick backups floating IPs for high availability and managed firewalls are all simple to configure. While you can set these things up yourself on a dedicated server it requires more manual effort.

A Simple Framework for Choosing

So how do you decide? Ask yourself a few simple questions about your workload.

1. Is the workload stable and long lived? If you are running a database a background job processor or a primary application server that will be on for months or years a dedicated server is probably cheaper.

2. Does the workload need to scale up and down quickly? If you need to handle sudden traffic spikes by adding and removing machines on demand cloud servers are the clear winner.

3. Is performance the absolute top priority? For CPU or disk intensive tasks a dedicated server will almost always provide more power for your money.

Here is a simple table to summarize the comparison.

Feature	Cloud Server	Dedicated Auction Server
Pricing Model	Pay-per-hour	Fixed monthly
Scalability	High (API driven)	Low (Manual)
Performance	Good (Shared CPU)	Excellent (Dedicated CPU)
Best For	Web frontends, spiky traffic	Databases, build servers, stable loads
Setup Time	Seconds	Hours

The Hybrid Approach Is Often Best

You do not have to choose just one. The most effective and cost efficient setup often involves using both types of servers together. This hybrid approach lets you get the best of both worlds.

Use a powerful dedicated auction server for your stateful services. Your database is a perfect candidate. It needs high performance and benefits from a stable environment. A build server is another great fit.

Then use cheaper flexible cloud servers for your stateless web frontends. Put them behind a load balancer. If you need to scale you can add or remove these web servers without touching your critical database server.

Hetzner makes this easy with their free vSwitch feature. You can create a private network that connects your cloud and dedicated servers. They can then communicate over a secure private IP address range with very low latency. This is the key to making the hybrid model work seamlessly.

# Example docker-compose for a web app
# that connects to a database on a dedicated server
services:
  web:
    image: my-web-app
    ports:
      - “80:8000”
    environment:
      - DATABASE_URL=postgres://user:pass@10.0.0.2:5432/mydb
# Note: 10.0.0.2 is the private IP of the dedicated server
# connected via Hetzner vSwitch.

By carefully considering the nature of your workload you can stop overpaying for flexibility you do not need. Look at your services. Identify the stable resource hungry ones and consider moving them to a cheaper more powerful dedicated machine. Your wallet will thank you.

— Rishi Banerjee
September 2025