Hetzner and Kamatera are both unmanaged VPS providers without managed services ambitions. Both are significantly cheaper than DigitalOcean. Both assume the operator is technically capable of running their own infrastructure.

The difference is in how each prices compute. Hetzner sells fixed-configuration instances — dedicated AMD EPYC vCPUs, predefined RAM and storage — at prices that lead EU independent cloud benchmarks. Kamatera sells components: CPU type and count, RAM, and storage priced independently, assembled by the operator into the exact configuration the workload requires.

The comparison resolves quickly: does the workload fit Hetzner's standard configurations without meaningful over-provisioning, and does the deployment fit inside four locations?

Hetzner tends to suit engineering teams whose deployment is EU-primary and whose workload fits standard configuration tiers — where dedicated AMD EPYC vCPUs at consistent performance and EU-leading pricing are the relevant criteria.

Kamatera tends to suit operators with non-standard resource profiles that map poorly to fixed tiers — high-RAM/low-CPU workloads, compute-burst applications, or deployments in Middle Eastern, African, or Southeast Asian geographies where Kamatera's data center coverage exceeds Hetzner's four locations.

For EU-primary deployments with standard resource profiles, Hetzner's price-per-core leads most cloud market comparisons. The case for Kamatera builds when the workload's resource ratio breaks standard tier assumptions, or when the geographic requirement falls outside Hetzner's four-location footprint.

Hetzner's product is maximum compute efficiency within its infrastructure model. Own data centers, own hardware, AMD EPYC dedicated vCPUs in the CCX series with no oversubscription — and pricing that consistently leads independent EU cloud benchmarks. The fixed-tier model bundles CPU, RAM, and storage into coherent configurations because the standard ratios serve most developer workloads adequately. The operator selects the tier; Hetzner's hardware delivers consistent performance.

Kamatera's model rejects fixed bundling. CPU type and count, RAM, and storage are independently priced components. The premise is that standard tiers impose waste on workloads that don't conform to standard ratios. A database server needing 16 GB RAM and 2 vCPU should not pay for the CPU a 4 vCPU fixed tier includes. An application with variable CPU requirements and low storage needs should not carry NVMe storage costs it won't use. Kamatera's value is precision — paying for exactly what the workload requires.

Kamatera's trade-off is that configurability requires the operator to know what they are configuring. The interface assumes infrastructure knowledge. The documentation is narrower than Hetzner's community-documented ecosystem. Misconfigured servers produce poor economics — cheap if assembled correctly, frustrating if the resource profile isn't understood before provisioning.

Hetzner operates four locations: Nuremberg, Falkenstein, Helsinki, and one US location in Virginia. For EU-primary deployments — especially those requiring German data residency under German law, including CLOUD Act exclusion — Hetzner's footprint provides adequate geographic positioning. Outside Europe and Virginia, Hetzner cannot serve the deployment.

Kamatera operates data centers in Europe, North America, Asia-Pacific, and the Middle East — including locations in Israel, Hong Kong, and the Netherlands, among others. For deployments requiring presence in geographies that Hetzner does not serve, Kamatera's footprint provides options that are rare among budget VPS providers. The Middle East coverage in particular exceeds most independent cloud competitors.

For EU-primary teams, the footprint comparison is largely irrelevant — both providers have adequate EU coverage. The footprint distinction matters for teams with global or regional requirements outside Europe and the US East Coast, where Hetzner's four locations create a hard geographic constraint.

Hetzner's CCX dedicated cloud series allocates specific AMD EPYC cores without oversubscription. Independent benchmarks consistently confirm performance-per-core that leads the EU independent cloud market. For CPU-intensive workloads — build pipelines, API services under concurrent load, data processing — Hetzner's dedicated vCPU model produces predictable, well-documented performance at pricing that is difficult to match.

Kamatera's performance is a direct function of the configuration the operator assembles. With accurate resource profiling, Kamatera can produce configurations that perform competitively for non-standard workload profiles. The risk is misconfiguration — under-allocated CPU for CPU-sensitive applications, or insufficient RAM for memory-intensive workloads. The infrastructure quality is adequate; the performance variable is the operator's configuration accuracy.

For standard workload profiles at EU locations, Hetzner's benchmark consistency and community documentation provide more predictable production behavior. For non-standard profiles where fixed-tier over-provisioning is significant, a correctly configured Kamatera server can produce better economics at equivalent performance — but that 'correctly configured' prerequisite carries weight.

Hetzner's EU pricing is among the lowest in any cloud market globally for dedicated vCPU configurations. A CCX32 — 4 dedicated AMD EPYC vCPUs, 16 GB RAM — at €16/month is a reference point that competitors struggle to match in European data centers. For standard workload profiles, Hetzner's fixed-tier pricing is the most economical option at the dedicated compute tier.

Kamatera's component pricing creates genuine cost efficiency for workloads with non-standard resource ratios. A high-RAM, low-CPU configuration assembled on Kamatera will typically cost less than the nearest Hetzner tier meeting the RAM requirement — because the Hetzner tier bundles dedicated vCPUs the workload does not need. Kamatera also offers hourly billing and a $100 free trial evaluation window, both of which reduce the cost of evaluating whether the component model fits the workload.

The total cost calculation favors Hetzner for standard resource profiles at EU locations — its fixed-tier pricing is simply lower. The calculation shifts toward Kamatera when the workload profile is non-standard and the over-provisioning at the nearest Hetzner tier is significant, or when the geographic requirement falls outside Hetzner's footprint.

You gain pricing precision with Kamatera. You give up compute efficiency at standard tiers. With Hetzner, the trade runs in reverse.

Hetzner aligns with engineering teams whose deployment is EU-primary and whose workload fits standard configuration tiers. You gain the best dedicated vCPU price-to-performance in the EU independent cloud market, consistent benchmark results, and a developer community with extensive documentation. You give up Kamatera's component pricing precision and geographic coverage beyond four locations.

Kamatera aligns with operators whose workload has a non-standard resource profile that fixed tiers cannot serve efficiently — or whose deployment requires regions outside Hetzner's four locations. You gain component-level pricing precision and geographic coverage in the Middle East, Asia-Pacific, and other geographies Hetzner does not reach. You give up Hetzner's EU price-to-performance leadership and community-documented benchmark consistency.

A practical diagnostic: identify the nearest Hetzner tier to the workload's actual resource requirements and calculate the over-provisioning cost. If that cost is negligible, Hetzner's EU pricing tends to align. If significant, or if the deployment requires a geography outside Hetzner's four locations, Kamatera's component model warrants evaluation.

The decisive question is whether the workload's resource profile fits Hetzner's standard tiers without meaningful waste — and whether the deployment fits inside four locations.

For EU-primary teams with standard workload profiles, Hetzner's dedicated vCPU pricing and benchmark consistency tend to be the most favorable unmanaged infrastructure available. The fixed-tier model is adequate, and the price-per-core advantage is significant.

For teams with non-standard resource profiles — disproportionate RAM, variable CPU, high-storage-low-compute — or with geographic requirements in the Middle East, Southeast Asia, or other regions Hetzner does not reach, Kamatera's component model aligns with those specific constraints. The operator configuration burden is the trade.

You gain EU compute efficiency and benchmark predictability with Hetzner. You give up configuration precision and geographic reach. With Kamatera, the trade runs in reverse.