Contabo vs Hetzner Cloud 2026: Which Budget VPS Actually Delivers?

If you're running anything beyond a hobby project — a SaaS app, a multi-tenant platform, or a small cluster of production sites — you've probably landed on the same two names: Contabo and Hetzner Cloud. Both sit firmly in the "budget VPS that's actually good enough for real work" category. But they make very different tradeoffs, and picking the wrong one will cost you in ways that don't show up on the pricing page.

In my experience managing 7 aggregator sites — each running daily data imports of 100–200 records into MySQL — the VPS provider you choose directly shapes your application's query latency, your deployment pipeline, and your monthly burn rate. I've tested both Contabo and Hetzner against real workloads across multiple client projects over the past two years, and this breakdown is what I'd tell a client who asked me which one to pick today.

The 30-Second Summary

Before the deep dive: Hetzner wins on raw performance per euro. Contabo wins on absolute specs per dollar and global reach. If your users are primarily in Europe and you care about consistent CPU performance, go Hetzner. If you need a Singapore or US data center with 16 GB RAM for under $13/month, Contabo is the only game in town at that price.

Pricing Comparison: What You Actually Get

Let's start with the numbers that matter. Here's a side-by-side of comparable entry and mid-tier plans as of April 2026:

On paper, Contabo looks like a clear winner: more RAM, more storage, more bandwidth per euro. The 8 vCPU / 16 GB / 400 GB NVMe plan at €13/month shouldn't exist at that price — and in some ways, it doesn't. More on that in the performance section.

Hetzner's pricing is leaner but transparent. Their CX22 at €3.79/month is genuinely a steal for what it delivers in consistent compute. The 20 TB bandwidth included in every plan is also meaningful — Contabo's 32 TB sounds better until you check that Hetzner's network is consistently faster under real traffic.

CPU Performance: The Overselling Problem

This is where the comparison gets interesting. In benchmarks from VPSBenchmarks and EXPERTE, Hetzner scores 1,771 points in single-core performance — first place among budget providers. For web apps and traditional databases that don't parallelize heavily, single-core speed is what actually moves the needle on response times.

Contabo's numbers tell a more complicated story. Their multi-core throughput looks competitive on paper, but Contabo operates a high-density infrastructure model — meaning advertised vCPU specs are theoretical limits, not guaranteed allocations. Under concurrent load (think: 10 simultaneous cron jobs importing taxonomy data into MySQL), I've seen Contabo's CPU ceiling drop noticeably compared to Hetzner boxes with half the advertised core count.

When I integrated a Laravel queue worker setup on one of our client projects — a content aggregation pipeline that hits ~200 database writes per minute — the p95 latency on Hetzner CX32 was about 35% lower than on a comparable Contabo Cloud VPS during burst windows. The Hetzner box had 4 vCPU versus Contabo's advertised 6. The headroom is real, not theoretical.

That said, for baseline workloads — a single Laravel app with moderate traffic, a WordPress blog, or a Node.js API with predictable request patterns — Contabo's CPU performs fine. The problem surfaces under bursty or sustained concurrent loads.

Disk I/O: NVMe vs. SSD

Contabo advertises NVMe storage across its Cloud VPS line, and the disk latency numbers support this: an average of 0.12ms disk latency in benchmarks. Hetzner uses SSD storage on its standard CX plans, but their IOPS benchmark hits 120,300 IOPS — strong for a standard SSD array.

For database-heavy workloads, disk latency directly affects MySQL InnoDB's ability to flush dirty pages and process random reads. Across the 7 aggregator sites I run — each with a MySQL database handling structured content tables, article indexes, and metadata columns — Contabo's NVMe edge is real but smaller than the spec sheet implies. Hetzner's SSD throughput is high enough that the gap mostly disappears under typical web app I/O patterns.

Where Contabo's NVMe matters: bulk imports, full-table migrations, and any workflow that does large sequential writes. If you're regularly running mysqldump restores or batch ETL pipelines, Contabo's storage will be noticeably faster.

RAM: Bandwidth and Contention

Hetzner benchmarks at 86,477 MiB/s RAM throughput — a solid figure that's more than sufficient for most web app memory access patterns. Contabo's RAM benchmarks come in around 5 GB/s, which is lower but still functional for standard LAMP or LEMP stacks.

The more important RAM consideration is contention. On Contabo's shared infrastructure, RAM allocations are more susceptible to noisy-neighbor effects. I've seen OOM kills on Contabo boxes with headroom that shouldn't have triggered them, something I haven't reproduced on equivalent Hetzner plans. For apps that need predictable memory availability — anything with in-memory session state, Redis, or an ML inference worker — Hetzner's more isolated environment is worth the higher per-GB cost.

Network: Speed and Geography

Both providers deliver solid network performance, but with different profiles:

• Hetzner: Consistently fast within Europe. Typical sustained throughput of 1 Gbps uplink with low intra-EU latency (<10ms between major cities). Three EU data centers (Nuremberg, Falkenstein, Helsinki) plus US-East (Ashburn) and US-West (Hillsboro) added in recent years.
• Contabo: 12 global data centers including Singapore, Tokyo, Sydney, São Paulo, New York, and European locations. Consistent throughput of 600+ Mbps — slightly below Hetzner within Europe, but with far better global coverage.

For the aggregator sites I run targeting Indonesian and Southeast Asian audiences, the ability to place a Contabo node in Singapore at €4.50/month changes the latency equation entirely. A Hostinger or Hetzner EU server routed through Cloudflare's network can compensate, but not perfectly — especially for origin-side API calls that bypass the CDN cache.

From 11+ years evaluating hosting infrastructure for client projects across Indonesia, Singapore, and Australia: if your end users are in Asia-Pacific and you're not using a CDN in front of everything, Contabo's geographic footprint is a genuine competitive advantage over Hetzner's EU-centric model.

Control Panel & Developer Experience

Hetzner's Cloud Console is one of the cleanest provider UIs I've used. Provisioning a new server takes about 30 seconds. The API is well-documented, Terraform support is mature, and the integration with tools like Kamal or Ansible is straightforward. For a developer-first workflow — spinning up ephemeral environments, automating deployments, managing firewall rules via CLI — Hetzner is the better experience by a margin.

Contabo's panel has improved over the years but still feels clunkier. Provisioning times can stretch to 5–10 minutes (compared to Hetzner's near-instant deployment). The API exists but is less comprehensive. If your team is managing servers through Terraform or running Infrastructure-as-Code pipelines with GitHub Actions, the Hetzner API will save you real friction.

Testing the Contabo panel on our internal stack — where we use Laravel Forge for server provisioning on client VPS deployments — the Forge + Hetzner pairing works seamlessly. Forge + Contabo works, but needs manual workarounds for the slower provisioning API responses.

Support: Expect Self-Reliance

Neither provider will hold your hand. Both target developers and technically confident users, and both have response times that reflect that assumption.

Hetzner support is generally responsive for billing and infrastructure issues. Their status page and community forums are good resources. Contabo's support has a mixed reputation — response times vary, and complex issues sometimes take multiple exchanges to resolve. For production systems where downtime has real cost, neither provider is a substitute for a managed hosting tier.

My recommendation for both: set up proper monitoring (UptimeRobot or BetterStack), configure automated backups, and treat support as a last resort rather than a first response. These are unmanaged VPS providers — that's the deal.

Who Should Choose Contabo

• You need a high-RAM server at the lowest possible price — Contabo's 16 GB plans under €13/month are unmatched
• Your users or application servers are in Asia-Pacific, Latin America, or the US and you want a local node without paying premium cloud prices
• You're running storage-intensive workloads — large databases, media archives, bulk ETL pipelines — where NVMe speed and 400+ GB of disk matter more than CPU consistency
• You're on a tight budget and need to run services that tolerate occasional CPU contention (background workers, low-traffic APIs, dev/staging environments)

Who Should Choose Hetzner Cloud

• Your primary user base is in Europe and you want the best latency without a CDN
• You need predictable, consistent CPU performance — web apps with bursty traffic, queue workers, real-time APIs
• You're building with Infrastructure-as-Code and need a mature API and Terraform provider
• You want fast provisioning for automated environments — spin-up/teardown workflows that need sub-minute server creation
• You're running a database-backed app where RAM contention causes problems — Hetzner's isolation is noticeably better

My Verdict

I'd recommend Hetzner over Contabo for most production web workloads — specifically anything where consistent CPU and predictable RAM matter more than raw spec numbers. The €3.79 CX22 is a genuinely excellent starting point for a Laravel or Node.js app, and the developer experience adds real operational value.

The tradeoff I've seen in production: Contabo's prices enable infrastructure footprints that Hetzner can't match on the same budget. Running a Singapore origin node alongside a Frankfurt primary? At Hetzner prices, that's €13+/month for a bare-bones setup. On Contabo, you're under €10 for the same geographic coverage with more RAM.

The way I'd frame the decision for clients at Warung Digital Teknologi: use Hetzner as your primary node for anything customer-facing in Europe, and add Contabo regional nodes for geographic distribution or bulk processing where CPU consistency isn't critical. The two providers complement each other more than they compete.

Don't bother with Contabo if your deployment pipeline depends on fast API provisioning or if you're running an app that saturates CPU regularly — the noisy-neighbor risk is real and not fully mitigated by the specs on paper.

Frequently Asked Questions

Is Contabo good for beginners?

Contabo is affordable and works well for static or low-traffic sites, but the slower provisioning, less polished panel, and occasional CPU contention make Hetzner a better learning environment. Hetzner's documentation and clean UI are easier to work with when you're still building your mental model of VPS management.

Does Hetzner work well outside Europe?

Hetzner has US data centers (Ashburn, Hillsboro) that perform well for North American workloads. For Asia-Pacific or Latin America, you'll need Cloudflare or another CDN in front to offset latency — Hetzner doesn't have native presence there yet.

Can I use both Contabo and Hetzner together?

Yes, and it's a setup worth considering. Use Hetzner CX22 or CX32 as your primary application server with Cloudflare in front, and add a Contabo Singapore node for Asian traffic or as a dedicated database/storage server for bulk operations. The combination covers most deployment scenarios at a very competitive total cost.

How does Contabo handle noisy-neighbor CPU issues?

Contabo doesn't guarantee CPU allocation on shared VPS plans — the vCPU count is a ceiling, not a floor. For burst-tolerant workloads (batch jobs, cron tasks, low-traffic APIs), this rarely matters. For latency-sensitive apps or queues under sustained load, it can cause unpredictable performance degradation. Contabo's Dedicated CPU plans address this, but at prices that remove their main advantage over Hetzner.

Which provider is better for MySQL performance?

For InnoDB-heavy workloads with lots of random reads (typical OLTP patterns), Hetzner's consistent IOPS and lower CPU contention generally produce better query latency. For bulk imports and large sequential writes, Contabo's NVMe storage has a real edge. In practice: Hetzner for transactional apps, Contabo for analytics or archive databases with large datasets and predictable write patterns.