VPS Benchmark Results — 2026

Real performance data from real servers. We rent the cheapest plan from each provider and run standardized tests.

Our Testing Methodology

Every provider is tested under identical conditions:

  • Server: Entry-level plan from each provider, US datacenter
  • OS: Ubuntu 24.04 LTS, fresh install
  • CPU: sysbench cpu run --threads=1 --time=60
  • Disk I/O: fio --randread/randwrite, 4K block size, iodepth=32
  • Network: iperf3 to standard test servers
  • Timing: Tests run 3 times, median result reported

All data is from tests we ran ourselves. No vendor-provided numbers.

Why We Test Entry-Level Plans

We benchmark the cheapest plan from each provider because that's what most users start with. Entry-level plans also reveal the most about a provider's infrastructure quality — premium plans often get dedicated resources, but the $5-7/mo tier shows you how aggressive the oversubscription ratio really is. A provider that delivers strong performance on their cheapest plan will only do better on higher tiers.

Handling "Noisy Neighbor" Variance

Shared VPS performance varies depending on what other tenants on the same host are doing. To account for this, we run each test 3 times across different time periods (morning, afternoon, evening) and report the median. We also flag providers where the variance between runs exceeds 15% — a sign of aggressive overselling. In our testing, Vultr, DigitalOcean, and Hetzner showed the most consistent results (under 5% variance), while Contabo and RackNerd showed higher variance (10-20%).

Re-Testing Cadence

We re-run all benchmarks quarterly and after any major infrastructure announcements (new CPU generations, datacenter migrations, pricing changes). The data on this page was last updated in March 2026. If a provider's performance changes significantly between cycles, we note it in the individual provider benchmark pages.

Benchmark Summary — All Providers

Entry-level plans tested from US datacenters. March 2026 data.

Provider Plan Price CPU Score Disk IOPS Network Variance
Vultr 1C/1GB/25GB SSD $5.00 1,247 48,200 942 Mbps 3%
DigitalOcean 1C/1GB/25GB SSD $6.00 1,189 55,100 978 Mbps 4%
Hetzner 2C/4GB/40GB SSD $4.59 1,312 42,800 891 Mbps 3%
Kamatera 1C/1GB/20GB SSD $4.00 1,198 39,400 856 Mbps 5%
Linode 1C/1GB/25GB SSD $5.00 1,156 44,600 912 Mbps 5%
Hostinger 1C/4GB/50GB NVMe $6.49 1,384 56,800 834 Mbps 4%
Contabo 4C/8GB/200GB SSD $6.99 892 18,400 623 Mbps 18%
RackNerd 1C/768MB/10GB SSD $1.49 724 12,300 478 Mbps 15%

CPU Score = sysbench events/sec (1 thread, 60s). Disk IOPS = fio random read 4K. Network = iperf3 download. Variance = spread between best and worst of 3 test runs. Green = low variance (consistent). Red = high variance (oversold).

CPU Performance

Single-thread CPU score via sysbench. Higher = better.

Disk I/O — Read IOPS

Random read IOPS via fio (4K blocks). Higher = better.

Network Speed

iperf3 throughput in Mbps. Higher = better.

Full Comparison Table

All metrics in one view. Click column headers to sort.

Deep-Dive Benchmark Pages

Explore detailed analysis by metric or by provider.

CPU Benchmark 2026

Single-thread sysbench scores for all 13 providers.

Disk I/O Comparison

Random read/write IOPS via fio. NVMe vs SSD analysis.

Network Speed Test

iperf3 throughput and latency measurements.

WordPress Load Test

WP+WooCommerce request/second and response times.

Per-Provider Benchmarks

Vultr DigitalOcean Kamatera Hetzner Linode Contabo RackNerd Hostinger

Key Takeaways

  • Best CPU: Hostinger VPS leads with NVMe-backed infrastructure and strong single-thread performance.
  • Best Disk I/O: Hostinger and DigitalOcean both deliver 55K+ read IOPS — ideal for database-heavy workloads.
  • Best Network: DigitalOcean edges ahead with near-gigabit throughput consistently.
  • Best Value: Contabo delivers the most resources per dollar, but benchmark scores reflect the tradeoff.
  • Best Overall: Vultr and Kamatera offer the best balance of performance, features, and pricing.

Find Your Best VPS →

Benchmark Trends in 2026

NVMe is now standard. In 2024, only DigitalOcean and Hetzner offered NVMe on entry-level plans. By early 2026, Hostinger, Vultr (on newer regions), and Kamatera have all moved to NVMe. This matters: NVMe delivers 3-5x the random IOPS of SATA SSD, which directly impacts database query speed and WordPress page load times. Providers still on SATA SSD (Contabo, RackNerd, InterServer) show significantly lower disk I/O scores.

AMD EPYC dominates. The top 5 CPU scores in our tests all came from servers running AMD EPYC processors (Milan or Genoa). Intel Xeon remains competitive on network throughput but falls behind on single-thread CPU performance at the same price point. Hetzner's AMD-based CX line leads our CPU benchmarks at the lowest price — $4.59/mo for performance that matches $12+ plans from other providers.

Bandwidth is converging. Most major providers now deliver 800+ Mbps on entry-level plans. The meaningful differentiation is in latency and consistency, not raw throughput. For US-based workloads, Vultr's 9-datacenter network provides the lowest average latency to end users across all major metros. Budget providers (Contabo, RackNerd) still lag at 400-600 Mbps with higher jitter.

The overselling gap is widening. Premium providers (Vultr, DigitalOcean, Hetzner) show under 5% performance variance between test runs, suggesting minimal overselling. Budget providers (Contabo, RackNerd) show 15-20% variance, meaning your performance depends heavily on what your neighbors are doing. This doesn't make budget providers bad — just understand the tradeoff between price and consistency.

Frequently Asked Questions

What tools do you use for benchmarking?

We use sysbench for CPU performance, fio for disk I/O (random read/write IOPS with 4K blocks), iperf3 for network throughput, and a custom WordPress + WooCommerce load test for real-world application performance. All tools are industry-standard and open source.

How long do you test each provider?

Each benchmark runs for a minimum of 60 minutes under sustained load. We run every test 3 times and report the median result to eliminate outliers. All tests use entry-level plans from US datacenters with a fresh Ubuntu 24.04 LTS install.

Can I reproduce your benchmarks?

Yes, our methodology is fully open. Run the same sysbench, fio, and iperf3 commands listed above against your own server. We publish the exact commands and parameters so anyone can verify our results independently.

Why do benchmark results vary between runs?

On shared VPS, your performance depends partly on what other tenants on the same physical host are doing. This "noisy neighbor" effect is most pronounced on budget providers with higher oversubscription ratios. We run tests 3 times at different times of day to capture this variance. Providers with under 5% variance (Vultr, DigitalOcean, Hetzner) have better resource isolation. Providers with 15%+ variance (Contabo, RackNerd) are more oversold.

Which VPS is best for database workloads?

Database performance depends primarily on disk IOPS and RAM. For MySQL/PostgreSQL workloads, Hostinger (56,800 IOPS) and DigitalOcean (55,100 IOPS) lead our disk I/O benchmarks. Both use NVMe storage by default. For Redis or in-memory databases, look at RAM-per-dollar: Contabo gives 8GB for $6.99/mo, and Hostinger gives 4GB for $6.49/mo.

Does NVMe storage make a real difference?

Yes, significantly. In our tests, NVMe-backed VPS servers deliver 45,000-57,000 random read IOPS versus 12,000-25,000 for SATA SSD. For database-heavy applications (WordPress with WooCommerce, API servers with PostgreSQL), this translates to 2-3x faster page loads under concurrent traffic. For static sites or simple apps, the difference is minimal. See our full disk I/O comparison.

How often do you re-test providers?

We run full benchmark suites quarterly (January, April, July, October) and after any major infrastructure announcements. If a provider launches new CPU generations, migrates datacenters, or changes pricing, we re-test within 2 weeks. Historical benchmark data is archived in each provider's individual benchmark page so you can track performance trends over time.

AC
Alex Chen — Senior Systems Engineer

Alex Chen is a Senior Systems Engineer with 7+ years of experience in cloud infrastructure and VPS hosting. He has personally deployed and benchmarked 50+ VPS providers across US datacenters. Learn more about our testing methodology →