4 active regions around the world

Pick the region geographically closest to the majority of your users — the latency references on each card are measured from nearby capitals. When compliance applies (GDPR, PDPA), let regulation decide first.

The region is fixed after creation, but you can snapshot the server and restore it in another region — support will gladly walk you through the move.

None — pricing is identical in all 4 regions. Choose by geography and compliance, never by price.

Private networks are per-region today. Cross-region traffic flows over the public interfaces — most teams wire it with WireGuard, which takes minutes on full-root servers.

The cards above list reference latency from major cities. As a rule of thumb: same continent <40ms, transatlantic ~80ms, Europe–Asia ~160ms.

Yes — your disk, snapshots and backups live in the region you chose. Nothing is replicated elsewhere without your action.

Riyadh (RUH-1) and Dubai (DXB-1) are in active build-out. Subscribe on the cards above and you'll be the first to know — and the first able to deploy.

ISO 27001 is the international standard for information security management. All four regions run inside facilities certified to it.

No — IP addresses are anchored to the region they are announced from; remixing them would break routing and geolocation.

In-region private traffic runs at up to 40Gbps backbone capacity and is free — ideal for app-to-database wiring.

No — the 99.99% uptime SLA applies uniformly to every region.

Each region announces its IP space through several independent carriers. If one carrier degrades, routes shift to the others automatically — usually within seconds and invisible to your users.

Frankfurt, currently — GCC routes into it run roughly 80–110ms from Riyadh and Jeddah, and around 90–120ms from Dubai, which comfortably serves websites, panels and APIs. Helsinki trails it slightly from the Gulf. For latency-critical workloads aimed purely at Gulf users, the honest advice is to register for the Riyadh launch notification — purpose-built regional capacity beats any remote optimization.

Each region card on this page lists a test endpoint you can ping or traceroute from your own connection — measured numbers from your actual ISP beat any published average. For application-level reality, spin up the smallest VPS in a candidate region for one month, run your stack against production-like traffic, then resize or redeploy where the numbers point. The test server can be destroyed the same day.

Use the notification form on the region's card — email only, no account needed. You will receive exactly two messages per region: a pre-launch date announcement and the go-live notice with introductory details, after which the list is retired. Subscribers also get first access during the early-capacity window, which matters because new-region capacity sells in waves. No marketing follows; the list exists for one purpose.

Web hosting, email, VPS and VDS run in all four active regions. AI GPU capacity currently deploys in Frankfurt and Helsinki only — power density and cooling for A100/H100 nodes are built out there first, with other regions following demand. Region availability shows live on each product's order form, so what you can select is always what is actually in stock.