Fits your load
VPS Nano
1 vCPU, 2 GB DDR5, 20 GB NVMe, 1 TB traffic
₹890$9.49/mo
Deploy NanoAbout NanoA site, a shop, a web app or an API, sized from your traffic, with request floods filtered before they reach it. On the AMD EPYC 4584PX at up to 5.7 GHz, with DDR5 memory and NVMe storage, in Mumbai, Singapore and Limburg.
Where most sites start
Prices exclude GST and are the same in every region. The calculator below sizes your own traffic.
Fits your load
1 vCPU, 2 GB DDR5, 20 GB NVMe, 1 TB traffic
₹890$9.49/mo
Deploy NanoAbout NanoRoom to grow
2 vCPU, 4 GB DDR5, 40 GB NVMe, 2 TB traffic
₹1,780$18.99/mo
| Plan | Traffic a month | Visits its traffic carries, no CDN | With a CDN |
|---|---|---|---|
| Nano | 1 TB | 3.3 million | |
| Mini | 2 TB | 6.7 million | |
| Starter | 4 TB | 13 million | |
| Growth | 6 TB | 20 million | |
| Pro | 8 TB | 27 million | |
| Elite | 10 TB | 33 million | |
| Max | 14 TB | 47 million | |
| Titan | 18 TB | 60 million |
Each plan's monthly traffic divided by an average visit: 2.6 pages of 2,311 KB, about 6 MB. The page weight is the global median from HTTP Archive; your pages can be far lighter or heavier. With a CDN in front, the calculator assumes 5% of the bytes still leave your server. Traffic only: this is how far the allowance goes, not how many visitors a plan can serve, which depends on your code. The calculator also checks CPU and memory for your kind of site.
Apply the Web server preset (TCP 80, TCP 443) or add the two rules yourself, with Web as the service. The preset ships with no packet-rate ceiling; add one once you know your traffic. A server holds up to 15 rules.
On plain HTTP, Prism reads each request and looks for the three attacks the help center lists:request floods above what the site should serve, requests with no user agent at all, and requests hammering paths the site does not serve. HTTPS is encrypted, so on 443 it judges the connections instead: their handshake, their rate and their TCP fingerprint.
A new server starts on Allow: ports without a rule stay reachable and get the always-on protections. Once 80, 443 and the port you administer the server on have rules, set Default for unlisted ports to Deny all. The Firewall tab
Most sites wait on a request, a database or a disk far more than they compute. A vCPU is processor time on shared cores, which is what a burst needs and the cheapest way to buy it. A production web app with its database on the same box.
A shop running imports, search and queue workers around the clock keeps its cores busy. On a VDS those cores are allocated to your server and are not scheduled for anyone else. One latency-sensitive service that should never wait for a shared core.
vCPU and traffic change while the server keeps running. Starter to Growth, as an example.
Memory and NVMe take effect when the server boots again. Restart from the panel when it suits you.
nginx, Apache or Caddy on Linux, IIS on Windows Server; PHP, Node, Python, Go or Java behind it. Nothing is preinstalled that you did not choose.
We look after the machine, the network and the filter. The software you install, its security and your backups are yours, and support can help you find your way around the panel and the platform.
No. It is a VPS with Full root / Administrator access and a clean operating system. You install the web server, PHP or your runtime, the database and the site, and you are responsible for the software on it and its security, as the Terms of Service set out.
No. Keep your domain where it is registered and point its DNS records at your server's IP address, which you will find on the server's overview in the panel.
Each plan has a monthly allowance, from 1 TB on Nano to 18 TB on Titan and 30 TB on UltraCore, on a 1 Gbit/s port. If a server uses more than its allowance, we may limit its port speed until the next billing period begins.
Yes, or any other PHP, Node, Python, Go or Java application. The calculator on this page has a setting for each, and the help center has guides for nginx and free TLS certificates.
Floods too large for the link are absorbed by 17 Tbps of upstream mitigation capacity. What arrives meets Clouduxe Prism on the host, which checks each packet against your port rules. On plain HTTP it also looks for request floods, requests with no user agent and requests to paths the site does not serve; HTTPS is encrypted, so there it judges the connections rather than the requests.
The one closest to most of your visitors: Mumbai for India, Singapore for Southeast Asia, Limburg for Europe. Hardware, protection and prices are the same in Mumbai, Singapore and Limburg.
Scheduled snapshots are included only where your plan says so, and support restores them on request. Keep your own copies of the files and the database off the server as well, and test that you can restore them.
Yes, from the panel. CPU and traffic change straight away; memory and disk take effect at the next restart. Nothing is reinstalled and your data stays where it is.
3 vCPU, 6 GB and 4 TB of traffic a month, with 17 Tbps upstream mitigation and Prism on every plan. Support runs 24/7/365.