Hong Kong NAT VPS
Real inventory — we sell only what we have

NAT plans share a public IP and expose SSH plus service ports through port forwarding, which is far more cost effective; dedicated IPv4 plans include their own public IP and suit services that need every port.

Sellable quantity = the sum of real node capacity configured in admin − sold − reserved by unpaid orders. Only enabled nodes with status online take part in sales, and unhealthy nodes are delisted automatically. The frontend receives change events over an SSE long connection with a 15-second polling fallback; if a plan shows “Sold out”, admin has not configured any sellable capacity for it — inventory is never hidden.

With a real Proxmox VE cluster connected, the instance enters the deployment queue after payment; when it finishes, the order detail page shows the SSH port, account and initial password and an in-site message is pushed. If the cluster is not connected or automatic provisioning fails, the order is marked as awaiting manual provisioning and your slot is kept — no specific timeframe is promised in that case.

Instance passwords are stored with AES-256-GCM encryption and are decrypted server-side only after ownership of the order has been verified; please change the password immediately after your first login.

“NAT shared IP” means several instances share one public IPv4 address and services are told apart by port: your SSH and each service port get a different public port assigned by the system, so the port number always has to be included — for example ssh -p 22001, or http://public-ip:28080. Dedicated IPv4 plans ship with their own public address and can use default ports, but cost more.

Public ports are assigned by the system when the instance is provisioned and cannot be picked by hand; the port inside your instance is yours to choose. On the order detail page, map a public port to port 80 or 443 inside the instance to serve HTTP(S) — the URL still carries the port, as http://public-ip:your-port, and the web server and TLS certificate are things you install inside the instance yourself. SSH permanently occupies one of your public ports and points at port 22 inside.

Our nodes are in Hong Kong, so they generally sit outside the mainland China ICP filing process. Whether you may offer a particular service from this machine depends on where that service is aimed, the regulations that apply to it, and each platform's own requirements — we draw no legal conclusions here. Please verify before ordering and read the relevant clauses in the Terms of Service.

The methods available are exactly the ones shown on the checkout page (account balance included). Once payment succeeds, the system allocates your ports and provisions the instance automatically; some orders are routed through a manual review. When provisioning finishes, the order detail page shows the SSH host, port, account and initial password — change the password first, then build the environment you need.

Suspension, release and any data retention period after expiry are governed by the Terms of Service; this page promises no number of days. Make backups a habit: keep a copy of anything important locally or on another machine, and re-purchase or export what you need before the expiry date.

Yes. On the order detail page you can request a refund within 48 hours of payment, no reason required: the instance and ports under that order are reclaimed immediately, and the money goes back the way it was paid — orders paid from your balance are credited back to the balance in full, orders paid through a third-party channel are refunded manually on that channel to your own payment account. For orders older than 48 hours, the Terms of Service apply; write to the support e-mail or Telegram listed in the footer and describe your situation.

No. These are self-managed VPS: installing the system, building the environment, deploying your application and fixing your own code are on your side. Our support covers whether the service itself works — whether the instance boots, whether ports forward as agreed, whether the order status is correct. That is also one of the reasons the price can stay this low.

Good fit: you want scripts and sites online around the clock, you are comfortable solving problems from a command line, and “shared IP plus port-based access” is acceptable to you. Poor fit: you need a dedicated IP to satisfy third-party allowlists or qualification requirements, you host business-critical systems with commitments to the outside world, or you want someone else to operate the server. For the first case, pick a plan that includes a dedicated IPv4; for the others, a managed provider is the better choice.

Yes. We commit to at least 99.5% monthly availability per instance (roughly no more than 3.6 hours of unavailability in a calendar month). The measure is the instance being up and its port forwards working as specified; it excludes maintenance announced in advance, and anything caused by your own actions inside the instance, suspension after a plan expires, or upstream carrier failures. If a month clearly falls outside that band, contact us through the support e-mail or Telegram listed in the footer with your order number — we verify the actual downtime and propose how to handle it, within the Terms of Service.