VPS 1C-2G
A small website or a development box
- CPU1 core
- RAM2 GB
- Storage50 GB NVMe
- Bandwidth1 TB
- Dedicated IP1
- Extra storageup to +500 GB
- Max bandwidth3 TB
Choose NVMe VPS servers with root access in 45 countries. Start small, scale when you grow, and keep your server close to your users. This page is your complete guide to choosing and running a VPS.
8 plans from $19.99/mo. Simple monthly pricing. Choose the plan, then choose any location.
A small website or a development box
A WordPress site with its database
Busy blogs and small online shops
Stores, APIs and Docker stacks
Several sites on one server
Databases and heavier apps
Build servers and busy back ends
Large apps and many services at once
Prices in USD, same price on renewal. Every plan includes full root access, KVM virtualisation, a dedicated IPv4 address and always-on DDoS filtering. Where a plan allows it, extra storage is $0.28 per GB per month and extra bandwidth is $2 per TB per month. No setup fees. Payment options are on the payment methods page.
Every country has its own page with local details. Pick the one closest to your visitors.
Buying a VPS is easier when you know what to look for. This guide explains each part of a plan in plain language, shows how to size a server for common workloads, and lists the questions worth asking before you pay. Use it as a checklist, and use the location pages when you are ready to choose a country.
A VPS plan is a bundle: processor, memory, storage, network traffic, a location, an operating system and a set of management tools. A good plan balances these for the work you plan to do, and a good provider makes each part clear. Below we walk through them one at a time.
In the animated diagram below, plan sizes grow as the memory and CPU they offer grow.
Memory decides how much your software can hold in fast storage at once. Web servers, PHP workers, application runtimes, databases and caches all use it. When memory runs out, the system starts swapping to disk, and everything slows down. For most projects, memory is the first resource to upgrade.
In the animated diagram below, NVMe reads random data many times faster than older storage.
Every plan uses NVMe storage. It is much faster than traditional SSDs for random reads and writes, which is the pattern that databases, content management systems and logs produce. Size matters too: operating system, applications, logs, uploads and backups all take space. Leave at least 20 percent free so that updates and logs never fill the disk.
If you need more space than your plan includes, add block storage or use object storage for large files such as media and backups. Keep databases on NVMe and archives on cheaper storage.
In the animated diagram below, traffic moves between your server and visitors in both directions.
Every plan includes a monthly traffic allowance and a public IPv4 address. IPv6 is available in most locations. Look at the traffic allowance if you serve video, downloads or large files, and use a CDN to reduce the load on the server. For mail servers, make sure you can set reverse DNS so that your messages are trusted.
The quality of the network depends on the location. Well-connected hubs with large internet exchanges give shorter routes and steadier speeds. That is why we list the exchange or network hub on each location page.
In the animated diagram below, distance adds delay to every request, so a nearby server answers faster.
You can resize a plan in minutes, but moving to another country means building a new server and migrating data. Choose the location early and choose it carefully. Match it to your users, your partners and your legal needs. If your audience is spread across regions, consider two servers or a CDN.
Linux powers most websites and applications. Ubuntu and Debian are the most common choices, AlmaLinux and Rocky Linux suit people who prefer the Red Hat family, and all are free of licence costs. Windows Server is the choice when your software needs it, such as MetaTrader, certain business tools or emulators. A licence add-on applies, and Remote Desktop gives you the usual Windows experience.
If you are new to servers, Ubuntu LTS has the most tutorials and the largest community. If you only need Windows for one program, consider whether a Linux server could run the same task through an API or a container.
In the animated diagram below, a dashboard graph and meters show the health of a server.
A VPS should be easy to manage. Look for a console, reboot and reinstall options, backups, a firewall, DNS tools and clear usage graphs. Also check what support covers. Our plans are self-managed, which means you control the server, and our team helps with the platform, the network, billing and setup questions.
| Workload | Suggested plan | Notes |
|---|---|---|
| Learning Linux, small bots | 1 core, 1 to 2 GB RAM | Low memory is fine |
| Blog or brochure site | 1 core, 2 GB RAM (VPS 1C-2G) | Use page caching |
| WordPress business site | 1 to 2 cores, 4 GB RAM (VPS 1C-4G or 2C-4G) | Add object cache for busy sites |
| Online shop | 2 to 4 cores, 8 GB RAM (VPS 2C-8G or 4C-8G) | Keep checkout fast |
| Game server | 2 to 4 cores, Exclusive CPU (VPS EX 2C-4G or 4C-8G) | Single-thread speed matters |
| Database server | 4 to 8 cores, 16 GB RAM (VPS 4C-16G or 8C-16G) | RAM first, then CPU |
| Build and CI server | 8 to 16 cores (VPS 8C-16G or 16C-32G) | More cores finish jobs faster |
| Windows trading terminal | Windows VPS 2C-4G or larger | Choose location near the broker |
Good testing takes an hour. Deploy a small server in the location you are considering, install a simple web server and measure. Use ping and mtr from a few locations that matter to you, run a quick disk test with fio, check CPU with a short benchmark and load a test page from a browser on your phone. Compare the results with your current host or with a second location. Keep notes, because numbers make decisions easy.
In the animated diagram below, security works in layers from the network to your data.
Do the same few things on every server. Create a normal user, use SSH keys, disable password logins, enable the firewall, install automatic security updates, set up fail2ban and keep a backup of the clean system. Then install only what you need and remove what you do not use. A small attack surface is the best defence.
In the animated diagram below, backups on a timeline let you restore an earlier state.
Decide how much data you can afford to lose and how long you can afford to be offline. Those two numbers, often called RPO and RTO, tell you how often to back up and how fast to restore. Use backups for quick rollbacks, scheduled backups for routine recovery and off-server exports for disaster recovery. Document the steps so that anyone on your team can follow them in a hurry.
In the animated diagram below, a load balancer shares traffic across several web servers.
A single VPS is enough for many projects, and it is simple to run. When uptime becomes critical or traffic grows, add resilience: a second web server behind a load balancer, a database replica, shared storage for uploads and a monitoring system. Start with the part that fails first, usually the database or the single web server, and add one layer at a time.
We answer all of these on the plan, location and support pages, and we are happy to answer them in a ticket.
Hosting comes in several forms, and each fits a different need. Shared hosting is the cheapest and simplest, but you share resources and have little control. A VPS gives you your own virtual server and full control at a moderate price. A dedicated server is a whole physical machine for you alone, which is powerful but costs more and takes longer to set up. Cloud platforms with many managed services give you building blocks, but their pricing is complex and you pay for each piece.
A VPS sits in the sweet spot for most projects. It starts cheap, grows in steps, and you can create it in minutes. When your needs outgrow one server, you add more servers instead of changing your whole approach. If you later need physical isolation, you can move to a dedicated server, and the skills you learned on a VPS carry over.
In the animated diagram below, a hypervisor divides one host into isolated virtual servers.
Virtualisation divides one physical server into several isolated virtual ones. With KVM-style full virtualisation, each VPS runs its own kernel, so it behaves like a separate machine and can run Linux or Windows. Container-based VPS share one kernel, which is lighter but less flexible. Full virtualisation is preferred for production because it provides stronger isolation and lets you install custom kernels, firewalls and filesystems.
You do not need to manage the hypervisor. The platform handles the host machine, and you manage your server from the operating system upwards.
Managed hosting includes someone who installs and updates software for you. It is convenient but more expensive and less flexible. Self-managed hosting gives you the server and the freedom to run anything, and you are responsible for updates, security and backups. Our plans are self-managed. If you are new, start with a small server, follow a guide and learn the basics. Most people become comfortable within a few weeks.
These terms are often confused. Port speed is the maximum speed of the network connection, for example 1 Gbps. Bandwidth is the capacity at a moment in time. Traffic, or transfer, is the total amount of data moved in a month, for example 3 TB. Your plan includes a monthly traffic allowance. Typical websites use far less than the allowance, while video, downloads and backups can use much more. Check your usage in the client area and use a CDN for heavy files.
In the animated diagram below, the software stack sits in layers on top of the virtual server.
Most projects follow a simple path. First, everything runs on one server: web server, application and database. Next, the database moves to its own server so that each part can be sized for its job. Then you add a second web server and a load balancer, and a cache such as Redis. Finally you add monitoring, automated backups and a staging environment. You can stop at any step that meets your needs.
Imagine a small agency with five client sites. One VPS 2C-4G server (2 cores, 4 GB) hosts all five sites with page caching. A second small Cloud 1 server is used for staging and tests. They add one domain per client. The monthly bill is two servers plus domain renewals, and they can see each item. When one client grows, they move that site to its own server. This approach keeps costs low and gives each client room to grow.
In the animated diagram below, the buying checklist ticks off item by item.
In the animated diagram below, the web, application and database roles move onto separate servers connected privately.
A single server is fine until one of three things happens: the database competes with the web server for memory, a deployment needs downtime that users notice, or a failure would cost more than a second server. At that point, split the roles and add monitoring. Splitting early costs a little more each month but makes problems easier to find, because each server has one job.
Now that you know what to look for, pick a location page that matches your audience and read its sizing and security sections. If you need Windows, see the Forex Windows VPS or Emulator Windows VPS pages. If you need a name for your project, register a domain and connect it with a single DNS record. If you still have questions, contact us with a short description of your project, and we will suggest a plan and a location that fit your needs and your budget.
In the animated diagram below, meters and a graph show how a server uses its processor, memory and disk.
A plan page lists vCPU, memory, storage and traffic. Each number tells you something about what the server can do. The vCPU count tells you how many tasks can run at once. Memory tells you how much data can be held in fast storage. Storage tells you how much space you have for files and databases. Traffic tells you how much data you may transfer each month.
Do not choose by one number alone. A server with many vCPUs and little memory will stall as soon as the memory fills up. A server with lots of memory but a slow disk will wait on every write. Compare the whole bundle against the work you plan to do, and use the meters in your client area after deployment to see which resource is closest to its limit.
In the animated diagram below, plan sizes grow step by step, showing how memory and CPU increase together.
When a server runs out of memory, the operating system starts moving data to the disk, which is called swapping. Swapping keeps the system alive but makes everything slow, because a disk is far slower than memory. If memory runs out completely, the system may stop a program to free space, and the program disappears without warning.
The cure is to measure and to plan. Watch memory over a week, remember that peaks matter more than averages, and leave a safety margin of at least 20 percent. Reduce the number of worker processes if you cannot add memory, and move to the next plan if the pattern repeats. A small upgrade is far cheaper than an outage.
In the animated diagram below, the layers of software sit on top of the operating system you choose.
The operating system is the base layer of the stack. Ubuntu LTS is popular because it has the largest amount of documentation and a predictable release schedule. Debian is stable and lean, a favourite for servers that should change slowly. AlmaLinux and Rocky Linux follow the Red Hat family and suit teams that already use those tools.
All of them are free, all of them run the same web servers and databases, and all of them can be reinstalled in minutes. Pick the one that matches your experience. If you are unsure, take Ubuntu LTS, follow its guides and learn the commands that carry over to every other distribution.
In the animated diagram below, the first-login steps tick off one at a time.
The first login is the most important one. Connect with the details from your client area, change any default password, create a normal user and add your public SSH key. Test the key from a second terminal before you disable password login, so that you do not lock yourself out.
After that, update the system, enable the firewall and install the tools you need. Keep your private key safe, protect it with a passphrase and never share it. If you work in a team, give each person their own key so that you can remove one person without affecting the others.
In the animated diagram below, attacks are stopped at a shield before they reach the server.
You can filter traffic in two places. A firewall sits outside your server and blocks unwanted traffic before it reaches the machine. A host firewall, such as UFW or nftables, runs inside the operating system and gives you finer control. Using both is the safest approach.
Start with a simple rule set: allow SSH from your own address, allow web ports from anywhere and deny everything else. Add rules for each service you install, and remove rules for services you remove. Review the rules every few months, because forgotten open ports are a common way in.
In the animated diagram below, a load balancer shares requests across three web servers in turn.
Horizontal scaling means adding servers instead of making one bigger. It works best when your application does not keep important state on a single server. Store sessions in a shared cache, keep uploads in shared storage and use a central database, and any server can answer any request.
A load balancer then spreads the work and removes failed servers from rotation. It also lets you update servers one at a time without downtime. This pattern takes planning, but it is the standard way to handle growth and to stay online when hardware fails.
In the animated diagram below, a live graph and three meters show processor, memory and disk.
You do not need to watch hundreds of numbers. Five matter most: CPU use, memory use, disk space, network traffic and response time. Together they tell you if a server is healthy, close to a limit or in trouble. Set alerts at levels that give you time to act, such as 80 percent of memory or 85 percent of disk.
Graph the numbers over time. A slow climb in disk use warns of a full disk next month. A sudden spike in CPU may mean an attack or a bug. A steady rise in response time may show a database that needs an index. Trends tell stories that single readings cannot.
In the animated diagram below, data moves from the old server to the new one while the old one keeps serving visitors.
Migrations go well when they are planned. List what you run, build the new server, copy the data, test and then switch. Keep the old server for a few days so that you can go back if needed. Lower the DNS time-to-live a day before the switch so that visitors move quickly.
Do the final data sync during a quiet period, put the site in read-only mode if the data changes often, and check logs on both servers after the switch. Write down what you learned so that the next migration is easier.
In the animated diagram below, backups on a timeline let you roll back to an earlier state.
A recovery plan is only real if you have tried it. Once a quarter, restore a backup or a backup to a new server and check that the application works. Time how long it takes, and note any steps that were missing from your instructions.
Decide how much data you can lose and how long you can be offline, then match your backups to those numbers. Frequent backups reduce data loss, and automation reduces recovery time. Keep at least one copy away from the server and the location.
In the animated diagram below, a browser asks DNS for your domain and receives your server address.
A server needs a name that people can remember. Register a domain, create an A record that points to the server IP and add records for www and any other service. Use a low time-to-live when you expect changes and a longer one when things are stable.
You can keep several domains on one server and several servers behind one domain, using subdomains and virtual hosts. Keep your DNS records documented and review them when you add or remove servers. See the domain pages for the full guide.
In the animated diagram below, bars compare costs for different options side by side.
As the number of servers grows, so does the bill. Keep a list with the purpose, size and renewal date of each server. Remove servers that nobody uses, shrink servers that are oversized and merge small projects where it is safe to do so.
Use a CDN and caching to reduce load, so that you can choose smaller plans. Choose the Exclusive CPU family only where steady performance matters, and keep Cloud plans for everything else. Review the list every quarter and you will rarely overpay.
Pick a location, read a guide or add a domain.
A VPS handles many workloads. Here are the most common searches.
Pick the RAM, CPU and disk that match your workload.
Shared hosting puts many sites on one server and one set of resources. A VPS gives you your own virtual server with its own operating system, root access and fixed resources.
KVM is full virtualisation, so each VPS has its own kernel and strong isolation. It supports Linux and Windows guests and suits production workloads.
NVMe is a fast type of solid-state storage that connects directly to the processor bus. It reads and writes much faster than older SATA SSDs, which helps databases, caches and builds.
You can choose from 45 countries, from Germany and the United States to Pakistan, Brazil, South Africa and Australia.
Start with 2 vCPU and 4 GB of RAM for a typical website or small app, 1 vCPU and 2 GB for a bot or light tool, and 4 vCPU and 8 GB or more for databases and busy shops. You can upgrade later.
Cloud uses high-performance shared vCPUs and is the best value. Exclusive CPU reserves vCPUs for your server and gives steady performance under constant load.
Plans are self-managed, so you control the server. Our guides show you how to install, secure and monitor your stack.
Yes. Windows Server is available with a licence add-on, and we have dedicated pages for Forex and emulator workloads.
Upgrade the plan from the client area. Take a backup first as a safety net.
Yes. With Nginx or Apache virtual hosts you can host many sites on one server as long as you have enough memory and CPU.
VPS CC plans are the Cloud Compute plans. They have fixed disk and transfer, optional automated backups priced per plan, and the smallest one starts at $14.99 per month.
VPS HF plans are the High Frequency plans, with faster CPUs and fixed disk and transfer. They sit in the Exclusive CPU tab and have optional automated backups priced per plan.
VPS EX plans are in the Exclusive CPU tab and follow the standard plan pattern of cores, memory, 50 GB NVMe and 1 TB of bandwidth, at a higher price for reserved performance.
On the standard plans you can add storage at $0.28 per GB per month, up to the limit shown on the plan card.
On the standard plans you can add bandwidth at $2 per extra TB per month, up to the limit shown on the plan card.
Ubuntu Server 20.04, 22.04 and 24.04, Debian 11, 12 and 13, AlmaLinux 8 and 9, Rocky Linux 8 and 9, Fedora Server, Alpine Linux and Windows Server 2019 and 2022. A custom ISO can be uploaded.
Servers are usually online in about 60 seconds after payment.
No snapshots or backups are included on most plans. The VPS CC and VPS HF plans offer optional automated backups at a price shown on each plan card.
Pick a plan and a location and be online in minutes.