VPS hosting for self-hosted AI
Benchmarked VPS comparisons for n8n, Ollama & Supabase.
Real, reproducible methodology instead of copywriter reviews. Find the cheapest server that actually runs your self-hosted AI and automation stack — and size it in seconds with the cost calculator.
Providers we compare
| Provider | Best for | Affiliate |
|---|---|---|
| Hetzner | Best raw price-to-performance. Listed editorially — no affiliate link. | editorial |
| Contabo | Most RAM/storage per dollar; flat affiliate bounty per signup. | affiliate |
| DigitalOcean | Strong docs, dashboard, and brand trust; recurring affiliate commission. | affiliate |
| Vultr | Many regions and hourly billing; per-referral bonus up to $100. | affiliate |
Latest guides
-
Best VPS for n8n in 2026: Hetzner vs Contabo vs DigitalOcean vs Vultr
Which cheap VPS actually runs n8n well? We size n8n's real resource needs and compare the four hosts on price, RAM, and value.
best-vps -
Ollama VPS Requirements: RAM Needed for Llama 3 8B, 13B & 70B (CPU Inference)
How much RAM and CPU you need to run Ollama models on a VPS without a GPU, and which plans fit each model size.
spec -
How to Self-Host n8n on a VPS with Docker Compose (2026)
A clean, copy-paste Docker Compose setup to run n8n on any VPS with Postgres and HTTPS, in about 15 minutes.
tutorial