Alternatives to Hetzner.

Hetzner is the opposite case from most tools on this site: it is not a US incumbent people want to leave, it is one of the strongest European picks in our own directory. Family-founded in Gunzenhausen in 1997, EU-owned, EU-hosted in Falkenstein, Nuremberg, and Helsinki, ISO 27001 and BSI C5 Type 2 attested, with CLOUD Act exposure: None and prices starting at €4. When people search for a Hetzner alternative, the reason is almost never sovereignty. It is what Hetzner deliberately does not sell: managed databases, managed Kubernetes, enterprise SLAs, broad GPU capacity, or a French public-sector qualification.

This page maps eleven European alternatives against the same four questions we ask everywhere: who owns the company, where the data sits, what the certifications actually cover, and what the move costs. The useful framing is not "escape Hetzner" but "keep the sovereignty posture, add the services Hetzner leaves out, and know exactly what that premium buys."

The three names worth knowing up front are Scaleway (France, Paris; the closest thing to a feature superset of Hetzner, with managed Kubernetes, managed databases, serverless, and GPU instances), OVHcloud (France, Roubaix; the largest European cloud, strong bare-metal range and SecNumCloud-qualified tiers), and IONOS (Germany, Frankfurt and Berlin; the German peer with enterprise SLAs, BSI C5 and IT-Grundschutz). All three are EU-owned and EU-hosted.

Moving between European IaaS providers is the gentlest migration category on this site. Carried over:

• The sovereignty posture. Nine of the eleven alternatives are EU-owned and EU-hosted; Exoscale (Switzerland) sits just outside the EU with Swiss residency, and Contabo is EU-operated but US-PE-owned since 2022.
• ISO 27001 across the board, with BSI C5 on the German providers (IONOS, STACKIT, T Cloud Public) and SecNumCloud on the French ones (OVHcloud, Scaleway).
• Standard Linux images, cloud-init, SSH-key workflows, hourly or per-second billing.
• Terraform / OpenTofu providers everywhere on this page.
• S3-compatible object storage on Scaleway, OVHcloud, IONOS, UpCloud, and others.

Left behind, or paid for differently:

• Hetzner's price level. Nothing on this page matches it like for like. The honest budget range is 1.5 to 3 times your Hetzner bill, partly offset if managed databases let you drop self-run infrastructure and the engineer hours that go with it.
• The 20 TB included egress per server. Most alternatives meter traffic sooner; price your real egress before moving.
• Snapshot portability: Hetzner snapshots cannot be exported, so the move happens at the application and data layer, not the image layer.
• The hcloud CLI and the community tooling around it (hcloud-cloud-controller-manager, csi-driver). The alternatives have their own equivalents, but your scripts get rewritten.
• Hetzner's dedicated-server auction market, a price-hunting instrument with no real equivalent elsewhere in Europe (OVHcloud's Eco range comes closest).

1. Write down why you are moving before you shop. "We need managed Postgres and a real SLA" leads to Scaleway, OVHcloud, IONOS, or STACKIT. "We need SecNumCloud on the contract" leads to OVHcloud or Scaleway only. "We need cheaper" leads nowhere; Hetzner is the floor. Teams that skip this step tend to migrate twice.
2. Get your Hetzner estate into code first. Snapshots do not leave Hetzner, so the migration unit is your provisioning code plus your data. If servers were built by hand, write the Terraform or Ansible against Hetzner first and verify it, then port the definitions. This turns an error-prone copy job into a re-deploy.
3. Move stateless workloads first, data last. Deploy application servers on the target provider, point them at the databases still running on Hetzner (both ends are in EU data centres, latency inside Europe is usually 5–25 ms), and verify behaviour under real traffic before touching state.
4. Replicate databases, do not dump-and-restore them. Postgres logical replication or MySQL GTID replication from the Hetzner primary to the new managed database gives you a continuously synced target and a cutover measured in seconds. This is also the moment you stop operating the database yourself, which for many teams is the entire point of the move.
5. Sync object storage with rclone. Hetzner Object Storage and Storage Boxes sync to any S3-compatible target with rclone sync in both directions. Run the first pass days ahead, the final delta during cutover.
6. Compare the egress line item explicitly. Take your last three months of Hetzner traffic graphs, apply the target provider's egress pricing, and put the number in the decision document. This is the most common unpleasant surprise in Hetzner migrations.
7. Keep the Hetzner servers running for a month after cutover. Hourly billing makes the safety net cheap. Downgrade to the smallest instance sizes that can still hold your data copies, watch the new provider's bill settle, then delete.