Best VPS for AI in 2026 — Contabo, Hostinger, Hetzner, Vultr & DigitalOcean Compared

Last Verified: September 2026  |  Author: Tom George

Data transparency: Infrastructure specs, pricing, and benchmark data in this article are sourced from the VPS providers themselves and from neutral third-party benchmark studies. Sources are credited at the bottom of this article. Plan prices were checked live on each provider's own pricing page on 28 September 2026 and are shown in the currency each page displayed. We earn commissions from some hosts we recommend — this never changes our verdict.

Why Your AI Stack Doesn’t Belong on Someone Else’s Server

Something changed quietly in 2025. Running AI tools on your own server stopped being a hobby project and became a legitimate cost optimisation strategy for developers, small businesses, and solo operators.

The numbers make the case clearly. A commercial AI stack combining ChatGPT Plus, GitHub Copilot, a vector database, and API costs runs $390–$690+ per month for a small team. A self-hosted stack running Ollama, n8n, and Open WebUI on an 8 GB VPS — anywhere from about €5.50 to $48 a month depending on the provider — covers most of the same ground — with your data staying on your own infrastructure, no per-token billing, and no usage caps.

The catch is that most “best VPS” guides are written for WordPress sites or simple web apps. Self-hosted AI tools have fundamentally different requirements — they are RAM-hungry, CPU-intensive, and sensitive to storage speeds in ways that a typical WordPress benchmark does not reveal.

This guide is written specifically for running AI workloads. We compare Contabo, Hostinger, Hetzner, Vultr, and DigitalOcean on the metrics that actually matter for Ollama, n8n, LocalAI, and AI agents — using live provider pricing, published benchmark data, and official infrastructure specifications.

How We Researched This Article

Sources used in this article:

• VPS benchmark data — independent benchmark comparisons published at aimultiple.com, comparing CPU events per dollar, disk I/O, and network performance for Hetzner and DigitalOcean specifically (their published test set does not include Vultr, Hostinger, or Contabo)
• First-party pricing pages — hetzner.com/cloud, vultr.com/pricing, digitalocean.com/pricing, hostinger.com/vps-hosting, and contabo.com/en/vps, read live on 28 September 2026
• Official infrastructure documentation — published by Contabo, Hetzner, Vultr, DigitalOcean, and Hostinger on their own sites
• Ollama and n8n official documentation — RAM and CPU requirements published at ollama.com and n8n.io
• Hostinger official VPS documentation — published at hostinger.com/vps-hosting and hostinger.com/tutorials/n8n-vps-requirements, covering KVM plan specs and pricing
• Independent cost analysis — published at use-apify.com, documenting real $12–$25/month self-hosted AI stack costs
• DigitalOcean official comparison documentation — published at digitalocean.com, covering infrastructure specs and positioning vs Hetzner and Vultr

What Self-Hosted AI Tools Actually Need From a VPS

Before comparing providers, let us be precise about what your VPS actually needs to handle.

The RAM problem. This is where most newcomers get caught out. VPS providers advertise entry plans starting at $4–$6/month — but those plans typically have 1–2 GB RAM. You cannot run Ollama on 2 GB RAM. The minimum recommended RAM for running a 7B parameter model like Llama 3.1 8B comfortably alongside n8n is 8 GB. Most advertised starting prices are for plans that are genuinely useless for this workload.

The three hardware requirements that actually matter:

1. RAM — The Non-Negotiable Minimum

RAM determines which AI models you can run and how many simultaneous requests you can handle. Here is the practical breakdown by model size:

2. CPU — Cores and Clock Speed Both Matter

Without a GPU — which most VPS plans do not include — all inference runs on CPU. CPU speed directly determines how fast your AI tools generate responses. For n8n automation workflows that call Ollama, slow CPU means slow workflow execution.

High-frequency compute instances — available on Vultr specifically — offer 3+ GHz clock speeds rather than the standard 2–2.5 GHz. For CPU-bound AI inference, this single factor can improve response times by 30–50%.

3. Storage — NVMe vs Standard SSD

AI models are large files. Llama 3.1 8B quantized is approximately 4.7 GB. Loading a model from storage into RAM happens every time Ollama starts or when memory pressure causes a model unload. On standard SSD this can take 30–60 seconds. On NVMe this drops to 5–10 seconds.

For development workflows where you switch between models frequently, this difference is significant. Most providers in this comparison offer NVMe storage — but not on every plan. Contabo’s entry Core line, Vultr’s Regular Performance line, and DigitalOcean’s Basic Droplets use standard SSD, while Hostinger and Hetzner list NVMe across the plans we checked.

The Five Providers — What They Actually Offer

Hetzner

Hetzner is a German cloud provider that built a devoted following among developers and self-hosters by offering more hardware for the money than US-based competitors. It is still a strong choice for European deployments with heavy bandwidth needs — but its pricing changed in 2026, and this section reflects what Hetzner’s own pricing pages show today.

What changed in 2026. Hetzner raised prices by 30–37% in April 2026 due to DRAM costs. As of 28 September 2026 (re-checked on 12 October 2026), its Cost-Optimized plans — the CX and Arm-based CAX lines, including the 4 vCPU / 8 GB CX33 listed at €8.99 — are all marked “not available,” and the page notes that the number of servers in that plan is limited.

Source: hetzner.com/cloud/cost-optimized, checked 28 September 2026; price-rise figure from independent analysis at bitdoze.com, February 2026

What you can actually order for 8 GB. The cheapest orderable 8 GB Hetzner plan is the Regular Performance CPX32 (4 vCPU / 8 GB / 160 GB NVMe) at €35.99/month in EU locations. In the US locations (Hillsboro and Ashburn) the 4 vCPU / 8 GB CPX31 is listed at €62.99/month. For 16 GB, the EU CPX42 is €69.99/month and the dedicated-vCPU CCX23 (4 vCPU / 16 GB) is €86.49/month.

Source: hetzner.com/cloud/regular-performance and /general-purpose, EUR view with VAT at 0%, checked 28 September 2026

Where Hetzner still wins. Hetzner includes 20 TB of traffic per month on EU plans (US plans include 1–5 TB), bills hourly, and posts very low CPU variance in independent benchmarks — see the Benchmark Data section below for how that data should be read against today’s prices. If you are deploying in Europe and need consistent CPU and cheap bandwidth, it remains a serious option.

Honest limitations. Hetzner’s data centres are in Germany, Finland, the US (Hillsboro, Oregon and Ashburn, Virginia), and Singapore — nothing in South America, Africa, or the Middle East. Support is email-based (Hetzner lists it as 24/7) rather than live chat. And there is no managed database service — you manage everything yourself.

Vultr

Vultr is a US-based cloud provider with 33 data centre locations worldwide — the largest global footprint of any provider in this comparison. Their High Frequency compute instances, running at 3+ GHz, are the best CPU option available for AI inference without a GPU.

High Frequency compute — the AI workload differentiator. Standard VPS instances typically run at 2.0–2.5 GHz. Vultr’s High Frequency instances use NVMe storage and 3+ GHz processors. For CPU-bound AI inference where every cycle counts, this is the most important differentiator Vultr offers. An inference task that takes 30 seconds on a standard 2 GHz VPS may take 18–20 seconds on a 3 GHz High Frequency instance — a meaningful difference for interactive workflows.

Global reach. With 33 regions including locations in APAC, Latin America, Africa, and the Middle East, Vultr is the best choice for teams with globally distributed users or for developers who need low-latency inference from multiple geographic locations.

Pricing reality. The High Frequency 2 vCPU / 4 GB RAM plan is $24/month, and the 3 vCPU / 8 GB High Frequency plan is $48/month (the AMD High Performance 4 vCPU / 8 GB plan is also $48, and the Regular Performance 4 vCPU / 8 GB plan is $40). Vultr also sells cheaper entry plans from $2.50/month (IPv6-only, 0.5 GB) and $5/month (1 GB) — too small for Ollama. For pure AI inference performance the high-frequency premium is justifiable; for cost-sensitive 8 GB builds, Contabo and Hostinger are far cheaper.

Source: vultr.com/pricing, checked 28 September 2026; Vultr notes that pricing can differ by data centre location

DigitalOcean

DigitalOcean has built its reputation on exceptional developer experience — documentation, tutorials, community, and managed services that reduce the operational burden of running production applications. For developers who are new to VPS hosting or who want to spend less time managing infrastructure, this matters.

The developer experience advantage. DigitalOcean’s documentation is widely regarded as the best in the industry. Their tutorials cover practically every self-hosted application setup including Ollama, n8n, and Docker-based AI stacks. For a developer setting up a self-hosted AI environment for the first time, the time saved by good documentation can easily justify higher server costs.

Managed services. DigitalOcean offers managed PostgreSQL, MySQL, Redis, and Kubernetes — services that Hetzner does not provide natively. If your AI workflow includes a vector database, managed Redis for caching, or Kubernetes for orchestrating multiple services, DigitalOcean’s managed offerings reduce operational complexity significantly.

The honest value problem. Independent benchmark data shows DigitalOcean returning approximately 71 CPU events per dollar per month — versus Hetzner’s 780. For pure compute value, DigitalOcean is not competitive with Hetzner or Vultr. Where DigitalOcean justifies its pricing is in managed services, documentation quality, and operational simplicity — not raw hardware.

Source: AIMultiple VPS benchmark (research.aimultiple.com/vps-benchmark, updated 12 August 2026; 4 vCPU / 8 GB shared plans, Hetzner at $18.59/month and DigitalOcean at $48/month at test time)

App Platform. DigitalOcean’s App Platform allows containerised application deployment without managing the underlying infrastructure. For developers who want to run n8n or other containerised AI tools without handling server configuration, this is genuinely useful — though it comes at a higher price than a raw Droplet.

Pricing. Basic Droplets start at $4/month (512 MiB) and reach 8 GiB / 4 vCPU at $48/month and 16 GiB / 8 vCPU at $96/month. Since 1 January 2026 Droplets bill per second (60-second or $0.01 minimum), which suits short experiments.

Source: digitalocean.com/pricing/droplets, checked 28 September 2026

Hostinger

Hostinger is a budget-friendly provider that competes on spec-to-price ratio rather than raw RAM volume. Its KVM VPS lineup runs on AMD EPYC processors with NVMe SSD storage across all plans, and the KVM 2 tier lands right at the 8 GB RAM floor this guide recommends for a 7B model plus n8n.

The value case. Hostinger’s KVM 2 plan includes 2 vCPU cores, 8 GB RAM, 100 GB NVMe SSD storage, and 8 TB bandwidth, priced at $8.99/month on the 24-month promotional term, renewing at $14.99/month (list price $24.49). That is a full self-hosted AI stack spec — CPU, RAM, and NVMe storage all included — at one of the lowest entry prices among the providers compared here, though the renewal rate is about two-thirds higher than the intro price.

Source: hostinger.com/vps-hosting, 24-month term as displayed, checked 28 September 2026

The trade-offs are real. Hostinger runs its own proprietary hPanel control interface rather than a standard panel or a bare Linux shell experience some developers prefer. Its data centres span nine countries — the US, Brazil, France, Germany, the UK, Lithuania, India, Indonesia, and Malaysia — which is wider than Hetzner’s but well short of Vultr’s 33 regions.

Contabo

Contabo is a German provider that competes almost entirely on hardware per euro. Its entry Core VPS — Cloud VPS 4 — gives you 4 vCPU cores, 8 GB RAM, and 100 GB SSD for €5.50/month at list price, or €4.40/month for the first 24 months.

Source: contabo.com/en/vps, EUR view, checked 28 September 2026

The value case. That makes Contabo the cheapest 8 GB plan in this comparison — roughly one-sixth of the cheapest orderable 8 GB plan at Hetzner (€5.50 vs €35.99). Stepping up, Cloud VPS 6 (6 vCPU / 12 GB) is €7.50/month and Cloud VPS 8 (8 vCPU / 24 GB) is €14.00/month at list price, which gives room for a second model or heavier n8n workloads.

Honest limitations. Contabo’s VPS lines run on shared hosts (it points buyers who need single-tenant CPU to its Cloud VDS line), Core storage is standard SSD rather than NVMe, and the Cloud VPS 4 network port is 200 Mbit/s — so pulling a 4.7 GB model takes roughly three minutes at best. Support runs through chat and tickets (Contabo cites a two-hour average response time, which we have not tested), and its 11 locations across 9 regions — including India, Singapore, Japan, and Australia — are far fewer than Vultr’s 33 (see our budget VPS guide for the full trade-off list). Contabo’s higher-priced Performance line (Cloud VPS Plus, AMD EPYC, NVMe) is a step up if storage speed matters — see our full Contabo review for every product line, and our Ollama on Contabo guide for the order flow.

Head-to-Head Comparison for AI Workloads

*Hetzner: the 4 vCPU / 8 GB CX33 is listed at €8.99/month but marked "not available" on Hetzner's pricing page; CPX32 is the cheapest 8 GB plan that could be ordered on 28 September 2026 (Cost-Optimized availability re-checked on 12 October 2026). *Contabo: "unlimited" traffic as advertised — see Contabo's terms for fair-use details. Prices are shown in each provider's own displayed currency and are not converted. Region counts, storage types, and support channels were checked on each provider's own pages on 28 September 2026. Contabo's support page still describes Cloud VPS 4 as "75 GB NVMe" while its pricing page lists 100 GB SSD; we use the pricing page.

Head-to-Head Matchups

Hetzner vs DigitalOcean

For the same 8 GB tier, DigitalOcean’s Basic Droplet is $48/month (4 vCPU) versus Hetzner’s €35.99 CPX32 (4 vCPU) — much closer than the gap the older benchmark data suggests. Hetzner still includes far more bandwidth in the EU (20 TB vs 5 TB), while DigitalOcean brings managed databases, App Platform, and better documentation.

Vultr vs DigitalOcean

At 8 GB the two are within $8 of each other: Vultr is $40 (Regular, 4 vCPU) to $48 (High Frequency, 3 vCPU), DigitalOcean is $48 (Basic, 4 vCPU). Vultr adds high-frequency CPUs and a much wider region list; DigitalOcean adds managed services and documentation. On GPUs, Vultr’s entry A16 is cheaper per hour, while DigitalOcean’s RTX 4000 Ada is the more inference-oriented entry card — see the GPU section below.

Contabo vs Hetzner

Contabo’s Cloud VPS 4 (4 vCPU / 8 GB / 100 GB SSD, €5.50) costs about one-sixth of Hetzner’s cheapest orderable 8 GB plan (CPX32, €35.99). What Hetzner buys you for the difference: NVMe, 20 TB of EU traffic, and lower CPU variance in benchmarks. If you can live with shared CPU and slower storage, Contabo is the far cheaper way to run a 7B model.

Vultr vs Hetzner

Hetzner covers Europe, the US, and Singapore; Vultr covers 33 regions across six continents and offers high-frequency CPUs and GPUs that Hetzner’s cloud line does not. For a US- or Asia-based team, Vultr’s region reach usually decides this; for a Europe-based team with heavy downloads, Hetzner’s bandwidth allowance still matters.

Benchmark Data — What Independent Tests Show

Independent benchmark testing in 2026 compared CPU events per dollar across VPS providers — Hetzner and DigitalOcean are directly covered by this test; Vultr, Hostinger, and Contabo are not part of this specific benchmark and are evaluated on their own published specs instead (see notes below). The results for the tested providers are stark:

Source: AIMultiple VPS benchmark, updated 12 August 2026 (Hetzner and DigitalOcean rows only; the study also tests IONOS, Linode, GoDaddy, and Liquid Web) — CPU events measured via sysbench on 4 vCPU / 8 GB shared plans, pricing at time of test. We re-read the figures against the live study on 28 September 2026. Vultr and Hostinger rows reflect provider-published specs, not this benchmark. Contabo was not tested.

Read the value figure with today’s prices in mind. The 780 events/$ figure divides measured CPU output by the price at the time of the test. Hetzner has since raised prices and its Cost-Optimized plans are currently listed as unavailable, so this number likely overstates what you would get per dollar today — we have not re-run the calculation ourselves.

The variance number matters. DigitalOcean’s standard deviation of 174.1 means CPU performance varies more between runs. AIMultiple’s reading is that low variance suggests dedicated or isolated cores rather than oversubscribed shared pools — an inference we have not tested ourselves. Hetzner’s standard deviation was 35.2. For AI inference where response time consistency matters for user experience, Hetzner’s low variance is a significant advantage.

If you’re moving from shared hosting to a VPS in the first place, our Shared Hosting to VPS — The Honest Upgrade Guide walks through what actually changes during that migration.

Our Recommendations by Use Case

Personal AI Stack — Cheapest 8 GB

Our pick: Contabo Cloud VPS 4

4 vCPU, 8 GB RAM, 100 GB SSD, €5.50/month at list price (€4.40/month for the first 24 months). Runs a 7B model such as Llama 3.1 8B alongside n8n with little headroom, and costs a fraction of any other 8 GB plan here. The trade-offs are shared CPU and SSD rather than NVMe. See Contabo Cloud VPS plans →

What you can run on this spec: Ollama with a 7B model, n8n for workflow automation, Open WebUI as a chat interface, and a little headroom for other services.

Personal AI Stack — Better Storage and Support

Our pick: Hostinger KVM 2

2 vCPU / 8 GB RAM / 100 GB NVMe SSD / 8 TB bandwidth at $8.99/month on the 24-month promo term, renewing at $14.99/month. That covers the 8 GB RAM floor this guide recommends for a 7B model plus n8n, with NVMe storage and 24/7 live chat. The trade-offs are the proprietary hPanel interface and fewer locations than Vultr. See Hostinger VPS plans →

Small Team — 2–10 Users, Moderate AI Usage

Our pick: Contabo Cloud VPS 8 on a budget, or Hetzner CCX23 / Vultr High Frequency 16 GB for consistent CPU

For 10 users with moderate AI usage, plan for 16 GB RAM or more and at least 4 vCPU; beyond 20 concurrent users, consider horizontal scaling with a load balancer. Contabo’s Cloud VPS 8 (8 vCPU / 24 GB) is €14.00/month at list price on shared CPU. If you need dedicated or high-frequency CPU, Hetzner’s CCX23 (4 vCPU / 16 GB, dedicated) is €86.49/month and Vultr’s High Frequency 4 vCPU / 16 GB is $96/month — justify the premium only if inference response time is critical. See Vultr High Frequency plans →

Europe-Based, Bandwidth-Heavy Workloads

Our pick: Hetzner CPX32 (or CX33 if it returns)

If your users and data are in Europe and you download models often, Hetzner’s 20 TB of EU traffic and low CPU variance still matter. Order the CPX32 (4 vCPU / 8 GB, €35.99/month) today; check whether the cheaper CX33 has returned before you buy, since Hetzner currently lists it as unavailable.

Developer Who Wants the Best Documentation and Managed Services

Our pick: DigitalOcean

Honest recommendation: if you are new to VPS hosting, DigitalOcean’s documentation quality is worth paying for. Their tutorials for Docker, Ollama, and n8n are comprehensive and well-maintained. Their managed PostgreSQL and Redis reduce operational burden if your AI stack needs databases. You pay a premium for this — but for developers who want to focus on building rather than maintaining infrastructure, it is a reasonable trade-off. Explore DigitalOcean →

Global Deployment — Users Across Multiple Regions

Our pick: Vultr

With 33 data centre locations, Vultr is the only provider in this comparison with genuine global coverage. For AI applications serving users across APAC, Latin America, and Africa, Vultr’s geographic reach is unmatched by Hetzner, DigitalOcean, Hostinger, or Contabo.

View Vultr global data centre locations →

Non-Technical — Want It To Just Work

Our pick: Cloudways on DigitalOcean

If managing a raw VPS sounds overwhelming — SSH configuration, firewall rules, Docker setup, SSL certificates — Cloudways sits on top of DigitalOcean infrastructure and handles the management layer for you. You get the underlying cloud quality with a control panel that does not require server administration knowledge. This is the most expensive option in this comparison but the lowest-friction path to a working AI hosting environment.

Explore Cloudways managed cloud hosting →

Ready to start your AI stack? Pick the plan that matches your use case above and you can be running Ollama and n8n the same day.

Start with Contabo Cloud VPS 4 →

Try Hostinger KVM with NVMe storage →

Do You Need a GPU? Live GPU Prices From Vultr and DigitalOcean

Most self-hosted AI on a VPS runs on CPU, but if you want fast inference on larger models, Vultr and DigitalOcean rent GPUs by the hour. The table shows the cheapest on-demand cards we found on each provider’s pricing page; monthly figures assume the card runs 24/7 for 730 hours, and both providers bill by usage, so a GPU that runs a few hours a day costs far less.

Source: vultr.com/pricing and digitalocean.com/pricing/gpu-droplets (DigitalOcean's new GPU pricing took effect 1 August 2026), checked 28 September 2026. Monthly figures are our own arithmetic (hourly price × 730 hours). Contabo's GPU VPS (contabo.com/en/gpu-vps, read 28 September 2026) is billed monthly rather than hourly, so it has no 730-hour estimate. The plan also includes 18 vCPUs, 96 GB of RAM, and 900 GB of NVMe in Europe or US Central; Contabo quotes €999 excluding VAT, and the €799.20 figure is the 24-month rate (contabo.com/blog, 21 September 2026).

For a personal or small-team Ollama and n8n stack, CPU-only on an 8 GB plan is still the better value. A GPU makes sense when you serve several concurrent users or need larger models that CPU cannot run acceptably.

Best VPS for AI Agents (OpenClaw and Similar)

AI agents such as OpenClaw often call a hosted model over an API instead of running the model on your server. In that case the VPS only runs the agent and its tools, so it needs much less RAM than a self-hosted LLM — our rule of thumb is a small plan of 2–4 GB, not the 8 GB floor above.

If your agent runs its model locally through Ollama, the 8 GB guidance earlier in this guide applies again. For step-by-step OpenClaw options across Hostinger and Contabo, see Best OpenClaw Hosting: Managed vs VPS Compared; for automation workflows, see Self-Hosted n8n Hosting.

What a Self-Hosted AI Stack Actually Costs

Setting Up Ollama in n8n — Step by Step

The screenshots below are published directly by Ollama on their official documentation. They show the exact steps to connect your self-hosted Ollama instance to n8n — the same process you follow on any VPS running these two tools together.

To make this concrete — here is the full stack that independent analysis in 2026 shows running on a $16–$25/month VPS (the table below shows what an 8 GB plan costs at each provider today):

Minimum VPS spec for this stack:

• 4 vCPU / 8 GB RAM / 80+ GB NVMe SSD
• Ubuntu 22.04 or 24.04 LTS
• Docker and Docker Compose installed
• Ports 11434 (Ollama), 5678 (n8n), 3000 (Open WebUI) open in firewall

Monthly cost comparison for this spec (live prices, 28 September 2026):

Security Considerations for Self-Hosted AI

Five security basics for every self-hosted AI VPS:

• Enable UFW firewall and explicitly deny ports 11434 and any other internal service ports from public access
• Use SSH key authentication — disable password authentication
• Keep your OS and Docker images updated — run apt update && apt upgrade weekly
• Use a reverse proxy (Nginx or Caddy) with SSL for any services you expose publicly
• Enable automatic security updates for critical packages

7 Questions to Ask Before Choosing a VPS for AI Workloads

Frequently Asked Questions

What is the minimum VPS spec to run Ollama and n8n together?

Minimum 4 vCPU / 8 GB RAM / 80 GB NVMe SSD. The 8 GB RAM is the hard floor — below this you will encounter memory errors when loading 7B parameter models alongside n8n. At this spec you can run Llama 3.1 8B or Mistral 7B comfortably for personal or low-concurrency use. For 10 concurrent users, upgrade to 8 vCPU / 16 GB RAM.

Can I run AI tools on a $5/month VPS?

For very small models — 3B parameters or less — a 4 GB RAM $5–$10/month plan can work. For Llama 3.1 8B or similar 7B models, 4 GB RAM is insufficient. You will hit memory errors. The practical minimum for useful AI workloads is an 8 GB RAM plan, which starts at €5.50/month at Contabo (Cloud VPS 4, €4.40 for the first 24 months) and $8.99/month at Hostinger (KVM 2, renewing at $14.99). The cheapest orderable 8 GB plan at Hetzner is €35.99/month.

Do I need a GPU VPS for self-hosted AI?

No — for personal or small team use. CPU inference on a modern VPS handles 7B models well enough for practical automation workflows. The cheapest on-demand GPUs at Vultr and DigitalOcean work out to roughly $340–$1,200+/month if left running 24/7, which eliminates most of the cost advantage of self-hosting. GPU makes sense for production applications serving many concurrent users or for fine-tuning models — not for typical n8n + Ollama automation stacks.

Is Hetzner reliable enough for production AI workloads?

Yes for most definitions of production. Hetzner offers a 99.9% uptime SLA, NVMe storage on the plans we checked, and consistent CPU performance with very low variance in independent benchmarks. The limitations are geographic — Europe, the US, and Singapore only — email-based support rather than live chat, and the fact that its cheapest Cost-Optimized plans are currently listed as unavailable. For production workloads with global users, Vultr’s broader geographic coverage is worth the premium.

Which VPS is best for AI agents like OpenClaw?

It depends on where the model runs. If the agent calls a hosted model API, a small 2–4 GB VPS is usually enough, and Hostinger’s or Contabo’s entry plans are the cheapest options. If it runs the model locally, use an 8 GB plan or larger — Contabo’s Cloud VPS 4 at €5.50/month is the cheapest 8 GB option, and our OpenClaw hosting guide compares the Hostinger and Contabo routes.

Is Contabo good for running Ollama?

For small quantized models, yes — its Cloud VPS 4 (4 vCPU / 8 GB / 100 GB SSD) is €5.50/month and runs a 7B model with little headroom. The trade-offs are shared CPU and standard SSD storage. See our step-by-step Ollama on Contabo guide.

Contabo vs Hetzner — which is better for self-hosted AI?

Contabo is far cheaper for an 8 GB plan (€5.50 vs €35.99 for Hetzner’s cheapest orderable 8 GB plan on 28 September 2026). Hetzner offers NVMe, 20 TB of EU traffic, and lower CPU variance. Choose Contabo for budget experiments and Hetzner when consistency and bandwidth matter.

What is the difference between Ollama and LocalAI?

Both run local LLM inference. Ollama is simpler — one command to pull and run a model, clean API, excellent n8n integration. LocalAI is more flexible — supports more model formats, more configuration options, better suited for advanced use cases. For most users starting with self-hosted AI, Ollama is the right choice. LocalAI makes sense if you need specific model formats or advanced configuration that Ollama does not support.

Can I run my existing content pipeline on a self-hosted VPS?

Yes — with caveats. A self-hosted Llama 3.1 8B on a good VPS produces quality comparable to GPT-3.5 for most tasks. For content generation, summarisation, classification, and structured data extraction, local models perform well. For complex reasoning, code generation, or tasks requiring frontier model quality, cloud APIs (Groq, Anthropic, OpenAI) still have an edge. A hybrid approach — local models for volume tasks, cloud APIs for quality-critical tasks — is often the most cost-effective architecture.

How do I transfer my data if I want to switch VPS providers?

VPS migration is straightforward compared to managed hosting migration. You SSH into your old server, compress your application data and n8n workflows, copy them to your new server using scp or rsync, and redeploy your Docker containers. Contabo, Hostinger, Hetzner, Vultr, and DigitalOcean all use standard Linux environments — there is no proprietary lock-in. The entire process typically takes 1–3 hours depending on data volume.

Our Bottom Line

Contabo — the cheapest way to get 8 GB of RAM today. Cloud VPS 4 is €5.50/month (€4.40 for the first 24 months) for 4 vCPU and 100 GB SSD. Shared CPU and standard SSD storage are the price of admission.

Hostinger — when you want NVMe storage, live chat support, and a budget price. KVM 2 is $8.99/month on the 24-month term, renewing at $14.99. Not the right choice if you need a standard control panel or broad global data centre coverage.

Hetzner — best for Europe-based workloads that need consistent CPU and cheap bandwidth (20 TB on EU plans). Its Cost-Optimized plans are currently listed as unavailable, so the cheapest orderable 8 GB plan is CPX32 at €35.99/month.

Vultr — when you need global data centre coverage, high-frequency CPUs, or GPUs by the hour. Pay the premium only if inference speed or region reach is genuinely critical.

DigitalOcean — when you are new to VPS hosting and value excellent documentation and managed services over raw compute value. Worth it if operational simplicity matters more than cost optimisation.

Cloudways on DigitalOcean — when you want cloud infrastructure quality without managing a server. The managed layer eliminates most operational overhead at a cost premium.

Related: Managed Node.js hosting for AI-built apps (GreenGeeks, Hostinger). Step-by-step Ollama deploy on Contabo's budget CPU VPS.  |  Self-hosted n8n hosting options.  |  OpenClaw hosting: managed vs VPS.  |  Best hosting for WooCommerce stores doing serious volume.  |  Switching hosts? Here is what to watch for.  |  Best hosting for Indian businesses.

Image Credits & Data Sources

Data Sources

1. CPU events per dollar benchmark, CPU variance data: AIMultiple VPS benchmark, updated 12 August 2026 — research.aimultiple.com/vps-benchmark (six providers including Hetzner and DigitalOcean; does not test Vultr, Hostinger, or Contabo)
2. Self-hosted AI stack cost analysis ($12–$25/month Ollama + n8n + Coolify), RAM requirements by model size, concurrent user scaling specs: use-apify.com — use-apify.com/blog/self-host-ollama-n8n-coolify-vps
3. Ollama and n8n integration, RAM requirements, port configuration, Docker setup: Ollama official documentation — ollama.com and n8n official documentation — n8n.io
4. Hostinger KVM VPS plan specs and pricing (24-month term, renewal rates): hostinger.com/vps-hosting, checked 28 September 2026; n8n deployment guidance: hostinger.com/tutorials/n8n-vps-requirements
5. Contabo Core VPS plan specs and pricing: contabo.com/en/vps, EUR view, checked 28 September 2026
6. Hetzner Cloud plan specs, prices, and availability status: hetzner.com/cloud/cost-optimized, /regular-performance, and /general-purpose, EUR view with VAT at 0%, checked 28 September 2026
7. DigitalOcean Droplet and GPU Droplet pricing: digitalocean.com/pricing/droplets and /pricing/gpu-droplets, checked 28 September 2026
8. Hetzner 2026 price increases (30–37%), bandwidth comparison, value assessment: independent analysis published at bitdoze.com, February 2026
9. DigitalOcean infrastructure positioning vs Hetzner and Vultr, managed services overview: DigitalOcean official documentation — digitalocean.com/resources
10. Vultr plan pricing (Regular, High Performance, High Frequency) and Cloud GPU pricing: vultr.com/pricing, checked 28 September 2026; High Frequency specifications and global data centre footprint: Vultr official documentation — vultr.com

See how Contabo's plans have changed over time in the Contabo pricing and plan changelog.