Bare-metal servers, VPS instances, and GPU compute, provisioned in minutes, billed hourly.
No tickets, no calls, no human in the loop unless you want one. Provision through the dashboard, the API, or our small CLI — whichever fits your workflow. Below is roughly what the CLI flow looks like.
Big cloud providers run their fleet from spreadsheets, optimised for shareholder value. We run ours from a workshop, optimised for not waking up at 4am. The difference shows up in three places: who answers the phone, what is in the price, and what happens when something breaks.
The price on the page is what you pay. Bandwidth is bundled to a generous cap, snapshots are included, ingress is free. No outbound traffic billed in attoseconds. We have done one pricing change in seven years — it went down.
Network ops support is staffed by engineers, not a tier-one queue. Email noc@ at 2am with a packet capture and someone with root access will reply with a packet capture. Median response time last quarter: nine minutes.
Every machine in the fleet is bought, racked, and maintained by us. The bare-metal SKU you book is a real chassis you can come and visit. We are not a thin reseller of AWS regions, and the difference is visible in the latencies, the steady pricing, and the absence of phantom noisy neighbours.
We publish capacity, latency to a major peer, and uptime for every region we run. This is a snapshot from the live board — the actual page updates every thirty seconds and is open to anyone, customers or not.
One bare-metal SKU, one VPS plan, one GPU box. The full catalogue is on the Fleet page — nine SKUs visible there, more on request. Every server in every class ships with the same support, the same pricing model, the same SLA.
In stock
VPS
In stock
GPU
Low stock
// new account — first 30 days of any single SKU are at 50% list price. See the full catalogue or request a custom SKU.