Vultr vs Hetzner: Why the "Cheaper" Choice Depends on a Map

Open two browser tabs. In one, pull up Hetzner's cloud pricing. In the other, Vultr. Stare at the numbers until the cognitive dissonance kicks in. Hetzner's CX22 at $4.59/mo: 2 vCPU, 4GB RAM, 40GB SSD, 20TB bandwidth. Vultr's $5/mo plan: 1 vCPU, 1GB RAM, 25GB SSD, 2TB bandwidth. Four times the RAM. Ten times the bandwidth. Forty-one cents cheaper. On every rational metric — price, specs, CPU benchmarks — Hetzner wins so decisively that writing a comparison article feels like comparing a Honda to a bicycle.

And yet Vultr has a larger US customer base. That is not irrational either. Open a third tab — a map of the United States. Mark Hetzner's two US datacenters: Ashburn, Virginia and Hillsboro, Oregon. Now mark Vultr's nine: New Jersey, Chicago, Dallas, Los Angeles, Seattle, Atlanta, Miami, Silicon Valley, Honolulu. The pricing advantage that looked absolute suddenly develops a condition: it depends on where your users are. A user in Atlanta hitting Hetzner's nearest server in Ashburn sees 15ms. That same user hitting Vultr's Atlanta datacenter sees 2ms. For a blog, nobody cares. For a real-time API or a game server, those 13 milliseconds compound across every single interaction. This entire comparison reduces to one question: does your application care about those 13 milliseconds? The answer determines everything else.

Head-to-Head Comparison Table

The first half of this table makes Hetzner look untouchable. The second half explains why Vultr still has a market. Pay attention to the bandwidth row — that 2TB vs 20TB gap catches people off guard on their first invoice.

Pricing: Hetzner's Absurd Advantage

I keep rechecking these numbers because they look like a data entry error. They are not.

Plan-by-Plan Breakdown

At the $20/mo price point where Vultr gives you 2 vCPU and 4GB RAM, Hetzner's $8.49/mo CX32 already includes 4 vCPU and 8GB RAM — double the machine for under half the price. Bandwidth: 20TB on every Hetzner plan versus 2-5TB on Vultr. If you run a bandwidth-heavy workload — media serving, software distribution, a high-traffic API — the bandwidth allocation alone can represent hundreds of dollars in avoided overage charges per year.

So why does anyone choose Vultr? Two reasons. First, Vultr offers $100 in free trial credit and Hetzner offers nothing. If you want zero-risk evaluation, Vultr lets you run real workloads for 14 days before spending a cent. Second, and far more consequential: Vultr's 9 US datacenter locations versus Hetzner's 2. For the detailed geographic analysis, keep reading — it is the section that turns this seemingly one-sided pricing comparison into a genuine decision.

One note on backup costs: Vultr's flat $1/mo backup pricing is simpler and cheaper at scale than Hetzner's 20% model. On a CX22 at $4.59, Hetzner's backup costs $0.92 — comparable. On a CX52 at $32.49, it costs $6.50 versus Vultr's still-flat $1. A small detail that adds up across a fleet. Compare more providers on our price comparison table.

Performance & Benchmarks

We benchmarked from comparable US East locations — Vultr's New Jersey and Hetzner's Ashburn, VA. The machines we tested: Vultr's 2 vCPU / 4GB at $20/mo and Hetzner's CX32 (4 vCPU / 8GB) at $8.49/mo. Yes, Hetzner's test machine costs less than half. That is the point.

CPU Performance

Hetzner: 4,300. Vultr: 4,100. A 5% per-core advantage for Hetzner, which runs AMD EPYC Milan across its newer fleet. Vultr rotates between AMD EPYC and older Intel Xeon depending on which physical host you land on. The variance introduces a small lottery element on Vultr that Hetzner avoids. But the real story is per-dollar CPU performance: Hetzner's CX32 delivers twice the cores, each running 5% faster, for 58% less money. The performance-per-dollar gap is not close. Check our full benchmark page for how both compare against the rest of the market.

Disk I/O

Hetzner: 52,000 read IOPS / 44,000 write. Vultr: 50,000 read / 40,000 write. Hetzner wins disk I/O too — a 4% read and 10% write advantage. Both use local NVMe storage. Hetzner's disk performance combined with its pricing means that even storage-intensive workloads like WordPress with WooCommerce run faster on Hetzner at a fraction of the cost. The disk gap is modest in absolute terms but meaningful when combined with the 2x RAM advantage that lets Hetzner cache more data in memory.

Network Speed

Hetzner: 960 Mbps. Vultr: 950 Mbps. Functionally identical for any real-world use case. The meaningful network difference is the bandwidth allowance: 20TB on every Hetzner plan versus 2-5TB on Vultr. If you are serving media, large downloads, or high-traffic APIs, that 10x bandwidth gap at entry level can add hundreds of dollars in Vultr overage charges annually.

US Datacenter Geography: Where the Map Overrides the Spreadsheet

This is the section that transforms a seemingly one-sided price comparison into an actual decision. If Hetzner's pricing advantage is the thesis, this section is the antithesis.

Vultr: 9 US Locations

• New Jersey (East Coast hub)
• Chicago (Midwest)
• Dallas (South Central)
• Los Angeles (West Coast)
• Seattle (Pacific Northwest)
• Atlanta (Southeast)
• Miami (Florida / Caribbean)
• Silicon Valley (California)
• Honolulu (Hawaii / Pacific)

Hetzner: 2 US Locations

• Ashburn, VA (US East)
• Hillsboro, OR (US West)

Nine versus two. Vultr covers the country; Hetzner covers the coasts. I measured the practical impact: average latency across all 50 US states is 18ms on Vultr and 31ms on Hetzner. That 13ms gap is invisible for casual browsing but real for performance-sensitive workloads.

Specific scenarios where the gap matters most:

• Dallas user to nearest server: Vultr Dallas = ~3ms. Hetzner Ashburn = ~35ms.
• Atlanta user: Vultr Atlanta = ~2ms. Hetzner Ashburn = ~15ms.
• Chicago user: Vultr Chicago = ~2ms. Hetzner Ashburn = ~20ms.
• Miami user: Vultr Miami = ~3ms. Hetzner Ashburn = ~25ms.

For static websites, content blogs, or anything behind a CDN like Cloudflare, these latency differences are neutralized at the edge. For real-time APIs, game servers, trading platforms, or any application where the server-to-user round trip shows up in your monitoring dashboards, datacenter proximity outweighs every pricing advantage on this page. Our US Datacenter Guide covers how to match datacenter locations to your specific user base geography.

Features: Where Vultr Fights Back

Windows VPS

Non-negotiable for some users: Vultr offers Windows Server images. Hetzner is strictly Linux-only. If you need .NET hosting, RDP access, or Windows-specific software, this comparison ends here — Vultr is the only option. See our Windows VPS comparison for more options.

Free Trial

Vultr's $100 free credit over 14 days lets you deploy in multiple US locations, run benchmarks, test latency from your actual user base, and evaluate the platform without financial risk. Hetzner requires a credit card on signup with no trial credits. At $4.59/mo the financial risk is trivial, but the psychological and organizational difference between "free trial" and "pay now" affects evaluation behavior more than people admit — especially in teams where procurement requires demonstrating a trial before committing budget.

One-Click Marketplace

Vultr's marketplace is broader: WordPress, Docker, cPanel, Minecraft, GitLab, Plesk, and dozens more. Hetzner's covers the essentials (Docker, Nextcloud, GitLab) but is smaller. Honestly, if you are comparing Vultr and Hetzner, you are probably deploying via SSH and a deployment tool, not clicking "Install WordPress." But the marketplace breadth reflects a broader investment in developer experience.

DDoS Protection

Both include DDoS protection, but with different depth. Vultr's automatic mitigation has a longer track record in the US market and more community-documented effectiveness. Hetzner's basic protection handles standard volumetric floods. For most workloads, both are adequate. For high-risk targets like game servers or crypto-adjacent projects, Vultr's battle-tested DDoS scrubbing provides more confidence.

Terraform & Infrastructure as Code

Here is an irony: the cheaper provider has the better IaC story. Hetzner's Terraform provider is official, HashiCorp-verified, and maintained by Hetzner's engineering team. It covers every resource type and ships updates within days of new feature launches. Vultr's Terraform provider is community-maintained — functional but occasionally lagging behind API changes.

For teams where terraform apply is the deployment mechanism and the Terraform state file is the source of truth, this difference matters. When your Terraform provider is community-maintained, a breaking API change can leave you waiting for a volunteer to push a fix. Hetzner's first-party commitment eliminates that risk. The hcloud CLI and official Go/Python SDKs reinforce the same message: Hetzner takes developer tooling seriously despite charging a fraction of the price. For Docker-based workflows, both integrate well, but Hetzner's official tooling gives it the edge.

Support Comparison

Vultr has live chat. Hetzner does not. At 2 AM when your production server is unreachable, that distinction outweighs every benchmark and pricing table on this page.

In my testing, Vultr's live chat connected me to a human within 4 minutes; ticket responses arrived in under 2 hours. Hetzner's tickets ranged from 3 hours to a full business day. When I asked a technical question about IPv6 routing, Hetzner's response was more detailed and technically precise than Vultr's — but it arrived 19 hours later. Vultr optimizes for speed; Hetzner optimizes for depth. Your priority determines which is "better."

Neither offers phone support. If phone support matters, see the Vultr vs Linode comparison — Linode is the only major indie provider with a phone line.

The Bandwidth Story: 20TB vs 2TB

This deserves its own section because it is the most overlooked number in the comparison table. Hetzner includes 20TB of bandwidth on every plan. Vultr's entry plan includes 2TB, scaling to 5TB at the $40/mo tier.

For a typical blog or low-traffic API, 2TB is fine. But the moment your application starts moving real data — media serving, software distribution, a popular API, a WordPress site with a traffic spike during a viral post — Vultr's overage charges at $0.01/GB start accumulating. Pushing 10TB in a month on Vultr's entry plan means $80 in overage charges on top of the $5 base. On Hetzner, that same traffic fits comfortably within the included allowance at $4.59 total.

This is not a theoretical advantage. I run a media archive that pushes 8-12TB per month. On Vultr, that would cost $60-100/mo in overage alone. On Hetzner's CX22, it costs $4.59 total. The bandwidth allocation is Hetzner's second-strongest advantage after the raw pricing, and it disproportionately benefits the exact workloads — media, downloads, high-traffic APIs — that tend to need the most resources.

Who Should Pick Which: Concrete Scenarios

Pick Hetzner If:

You are bootstrapping a startup and $4.59/mo buys what $20/mo buys elsewhere. At the early stage, 4x infrastructure efficiency is not about saving money — it is about extending runway. A real SaaS application (API, database, Redis, background workers) can run on a single CX32 at $8.49/mo. On Vultr, the equivalent specs cost $40/mo. That is $378/year saved per server.

Your Terraform repo is the source of truth. Hetzner's official provider and SDKs mean fewer surprises when you run terraform apply. The IaC story is genuinely best-in-class for a provider at this price point.

You are self-hosting Nextcloud, GitLab, or Matrix. Applications that need lots of RAM and bandwidth but not necessarily multi-region distribution. Hetzner's 4GB entry plan with 20TB bandwidth is purpose-built for self-hosted infrastructure.

Your users are on one coast. Ashburn serves the East; Hillsboro serves the West. If 80%+ of your traffic comes from one side of the country, you do not need nine datacenters. Save your money.

Pick Vultr If:

You are running game servers across the US. A Minecraft server in Dallas, another in Atlanta, a third in Chicago. Sub-20ms ping requires geographic proximity that Hetzner's two locations cannot provide. Vultr's nine locations cover the US comprehensively.

You need Windows VPS, FreeBSD, or custom ISOs. Hetzner is Linux-only. Hard constraint, not a preference.

Your application is a DDoS target. Game servers, crypto projects, e-commerce during peak seasons. Vultr's battle-tested DDoS mitigation provides more confidence than Hetzner's basic protection for high-risk workloads.

You want a risk-free evaluation. Vultr's $100 free trial lets you deploy, benchmark, and test latency without spending a dollar. Hetzner charges from day one.

Your user base spans the continental US. If your analytics show meaningful traffic from the Midwest, Southeast, and Southwest in addition to the coasts, Vultr's 9-location coverage eliminates the 15-35ms latency penalty that Hetzner's 2-location footprint imposes on those regions. Use our VPS calculator to estimate the right plan size for your workload.

Benchmark Chart

Visual comparison from our test runs. Vultr tested on the $20/mo plan; Hetzner on the $8.49/mo CX32. The CX32 costs 58% less and wins or ties every metric.

Final Verdict

I run personal projects on Hetzner because the math is undeniable. I run client projects on Vultr when the client's users are scattered across the US. Both decisions are correct, because they optimize for different constraints.

Hetzner is the better deal by every objective pricing metric. Four times the RAM, ten times the bandwidth, faster CPU per core, faster disk I/O, official Terraform support, and 50-79% lower prices at every tier. If you are a developer who values infrastructure efficiency, a startup extending runway, or anyone whose user base is concentrated on one coast, Hetzner is the most rational choice in cloud computing. I have never felt like I was compromising on anything except datacenter proximity to the US interior.

Vultr is what you choose when geography matters as much as price. Nine US datacenters mean you can put a server within 15 milliseconds of almost any American city. You also get Windows support, a $100 free trial, live chat support, and a one-click marketplace that Hetzner does not match. None of that is free — you pay a 2-5x premium for it. But when your SLA depends on sub-20ms latency to Dallas, or your compliance team requires US-interior infrastructure, or your game server needs to be within a state's borders, the premium is the price of admission.

The question I ask every person who brings me this comparison: can your application tolerate 31ms of average US latency instead of 18ms? If yes, save your money and go with Hetzner. If that number makes you nervous, Vultr is worth every extra cent. For a related comparison, see our Hetzner vs DigitalOcean breakdown.