Table of Contents
- What $3 buys
- The 2 GB comparison in US datacenters
- After you order
- How to order and size it: a step-by-step guide
- The first hour after provisioning (Linux checklist)
- Four common setups and how many slices they need
- Plans and pricing: the slice system
- Three US locations and the facilities behind them
- What "managed" means here
- The four rules you accept at $3
- What the missing snapshot and API cost, by buyer size
- Who should buy it, and who should not
- FAQ (11 questions)
- Sources
What $3 buys
InterServer sells cloud VPS in units it calls slices. One slice is $3 a month and comes with 1 core, 2 GB of memory, 40 GB of SSD storage, 2 TB of monthly transfer on a shared 10 Gbps port, one IPv4 address, root access and a choice of Linux images (September 2026, interserver.net/vps). KVM virtualization and free DDoS protection are included at every size.
Two things make that number unusual. First, 2 GB of RAM is the working minimum for a small production stack (a web server plus a database plus a cache), and most US providers price that tier at $10 or more. Second, the VPS page lists $3 as the monthly rate with no introductory qualifier attached, and the grid above it is linear: two slices cost exactly twice one slice.
What the number does not tell you is written in the provider's own FAQ, and it matters more than any spec: there is no trial, no refund, and no hourly option. We come back to that in the four rules.
The 2 GB comparison in US datacenters
This is the table the rest of the review hangs on. Every row is a plan you can actually order in a US location, at the price shown on the provider's own pricing page or public API on the same day (September 2026). Sorted by monthly price.
| Provider and plan | Monthly | vCPU | RAM | Storage | Transfer | Port | Billing |
|---|---|---|---|---|---|---|---|
| InterServer, 1 slice | $3.00 | 1 | 2 GB | 40 GB SSD | 2 TB | 10 Gbps shared | Monthly only |
| Contabo, Cloud VPS 4 | $6.60 | 4 | 8 GB | 100 GB SSD | Unlimited on a 200 Mbit/s port | 200 Mbit/s | Monthly, 24-month intro pricing |
| Vultr, vc2-1c-2gb | $10.00 | 1 | 2 GB | 55 GB SSD | 2 TB | Shared | Hourly or monthly |
| DigitalOcean, Basic 2 GB | $12.00 | 1 | 2 GB | 50 GB SSD | 2 TB | Shared | Per-second |
| Hetzner, CPX11 (Ashburn or Hillsboro) | $20.49 | 2 | 2 GB | 40 GB NVMe | 1 TB | Shared | Hourly or monthly |
Two notes on the table. Contabo wins on raw memory per dollar, but its entry plan runs on a 200 Mbit/s port, so it is a different kind of server; our Contabo review covers when that trade works. Hetzner's famous sub-$7 plans do not appear because they are not sold in US locations; in Ashburn and Hillsboro the entry plan is CPX11 at $20.49 with 1 TB of transfer (our Hetzner review has the full US and EU grids). RackNerd is absent because its cheapest offers are annual promotions rather than monthly plans; if you are willing to prepay a year, read that review before deciding.
Within the "2 GB, US, monthly" frame, InterServer is the lowest price by a factor of three. That is the whole case for it. The next three sections are about getting a working server out of it; the sections after that are about what the price leaves out.
How to order and size it: a step-by-step guide
The order form asks for five decisions. Taking them in this order avoids the two mistakes new buyers make most often: buying Windows with one slice (it needs two) and paying for a control panel they never open.
Step 1: pick the line
- Linux Cloud Compute (from $3): the default. KVM, root access, any of the Linux images or FreeBSD.
- Windows Cloud Compute (from $10): Windows Server on Hyper-V with Remote Desktop. Minimum 2 slices (1 core, 4 GB, 80 GB SSD, 4 TB), then $5 per additional slice. No control panels are offered on Windows.
- Storage Cloud Compute (from $3): the same slice price, but each slice carries 1 TB of SATA disk instead of 40 GB of SSD, with the same 2 GB RAM and 2 TB transfer. For backups, archives and file hosting, not for databases.
Step 2: pick the region
New York City (Secaucus, NJ), Los Angeles or Dallas. Choose by where your users are, not where you are: East Coast and European visitors are best served from New York, the Pacific and Asian traffic from Los Angeles, and Dallas splits the difference for a US-wide audience. If you need a region outside those three, this is not your provider; the US datacenter guide lists who covers which city.
Step 3: pick the slice count
Size to memory first; CPU follows at one core per two slices. The table in four common setups maps workloads to slice counts. If you are unsure, start at 1 or 2 and add slices later: the FAQ confirms slices can be added at any time, and the price is the same whether you buy them now or in month three.
Step 4: pick the operating system
- Ubuntu LTS: the safe default. Every tutorial on this site assumes it.
- AlmaLinux (CentOS-compatible): choose this if you intend to run cPanel or the Webuzo one-click stack; InterServer's own WordPress path is AlmaLinux with Webuzo.
- Debian: leaner base image, same package tooling as Ubuntu.
- FreeBSD: available; note that the control panel options are Linux-only.
- App Marketplace templates: preconfigured images (WordPress, Webuzo and others) attached to the order flow. Useful if you want the application installed before first login; skip if you prefer a clean OS.
Step 5: decide on a control panel, extra IPs and backups
- Panel or no panel. If you will manage one or two sites over SSH, skip the panel. If you host clients or want a web interface for domains, mail and databases, DirectAdmin Personal is listed at $3.75 a month (2 accounts, 20 domains); cPanel is licensed per account and priced on the order form; Plesk is also offered. Panels are ordered from My.InterServer after the VPS exists; cPanel installation on Cloud VPS is listed as free. Webuzo is the included-panel path: see Webuzo on an InterServer VPS. If you are buying a cPanel licence, see cPanel on an InterServer VPS.
- Extra IPs. One IPv4 is included; up to five more at $3 per IP per month.
- Backups. Nothing is included. Either add InterServer's Cloud Remote Backup Service or set up your own. The backup guide is the first-hour job, because InterServer does not advertise snapshots. Decide this before the server holds anything you would miss.
Payment is monthly by card; you can also load prepaid credit in the portal, but the FAQ is clear that prepaying earns no discount. There is no trial, so the first month is the trial.
The account side after checkout (My.InterServer, adding slices, panel licences, managed support) is the how to use an InterServer VPS guide. A site already running elsewhere is the migration guide.
The first hour after provisioning (Linux checklist)
You receive a root password and a public IP. On a 2 GB server the order of operations below matters: updates first, then a non-root user and key-only SSH, then the firewall, then swap, and only then the application. The commands are for Ubuntu 24.04; Debian is nearly identical.
1. Update and reboot
ssh root@YOUR_SERVER_IP
apt update && apt full-upgrade -y
reboot
2. Create a sudo user and move to SSH keys
adduser deploy
usermod -aG sudo deploy
# on your own machine, copy your public key up:
ssh-copy-id deploy@YOUR_SERVER_IP
# then on the server, disable password logins and root SSH:
sed -i 's/^#\?PasswordAuthentication .*/PasswordAuthentication no/' /etc/ssh/sshd_config
sed -i 's/^#\?PermitRootLogin .*/PermitRootLogin no/' /etc/ssh/sshd_config
# if /etc/ssh/sshd_config.d/ holds an override file, change it there too
systemctl restart ssh
Test the key login in a second terminal before you close the root session. If the key fails, you still have root to fix it.
3. Firewall
ufw allow OpenSSH
ufw allow 80/tcp
ufw allow 443/tcp
ufw enable
InterServer provides DDoS protection at the network edge; ufw is the host firewall and you still need it. Open only the ports your application uses.
4. Automatic security updates and brute-force protection
apt install -y unattended-upgrades fail2ban
dpkg-reconfigure -plow unattended-upgrades
systemctl enable --now fail2ban
5. Add swap (do this on every 2 GB server)
A 1-slice server has 2 GB of RAM and no swap by default on most images. Without swap, a memory spike kills the database instead of slowing it down. 2 GB of swap on the 40 GB disk is the standard trade.
fallocate -l 2G /swapfile
chmod 600 /swapfile
mkswap /swapfile && swapon /swapfile
echo '/swapfile none swap sw 0 0' >> /etc/fstab
sysctl vm.swappiness=10
echo 'vm.swappiness=10' >> /etc/sysctl.conf
6. Timezone, hostname, and a first backup job
timedatectl set-timezone America/New_York
hostnamectl set-hostname web1.example.com
# nightly database dump + off-server copy (adjust for your stack):
apt install -y restic
# restic init -r sftp:user@backup-host:/srv/backups # once
# then a cron entry, e.g. 0 3 * * * /usr/local/bin/nightly-backup.sh
The backup target can be the storage VPS line ($3 per 1 TB slice, in a different InterServer region), any S3-compatible bucket, or a machine you own. What matters is that it is not the same disk. The InterServer backup guide is the restic job written out. Our security hardening guide continues from here, and the WordPress on VPS guide covers the application layer if you are installing by hand instead of using Webuzo.
Four common setups and how many slices they need
| Workload | Slices | Monthly | Why this size | Notes |
|---|---|---|---|---|
| Static site, bot, VPN endpoint, uptime monitor | 1 | $3 | 2 GB is plenty; the 2 TB transfer covers a personal VPN comfortably | Add swap; see the VPN VPS page for WireGuard setup |
| WordPress or PHP site with a caching plugin | 2 | $6 | 4 GB leaves room for PHP-FPM workers, MariaDB and an object cache on one box | AlmaLinux + Webuzo template if you want one-click WordPress (Webuzo guide); Ubuntu + LEMP if you prefer to build it |
| WooCommerce, small app with Postgres, Docker host with 3 to 5 containers | 4 | $12 | 2 cores and 8 GB; the first size where CPU stops being the limit | Compare with Contabo Cloud VPS 4 ($6.60, 8 GB, slower port) before buying |
| Windows Server with RDP | 2 (minimum) | $10 | Windows needs the 4 GB floor; Hyper-V, RDP included | No control panels on Windows; see the Windows VPS page and the RDP guide |
| Production site you do not want to administer | 8+ | $24+ | The threshold where InterServer includes managed support for services, databases and panels | Still bring your own backups; the backup guide is the job |
Plans and pricing: the slice system
You choose any number of slices from 1 to 32 on the order form; every slice adds the same block of resources (September 2026):
| Slices | Linux | Windows | Cores | RAM | SSD | Transfer |
|---|---|---|---|---|---|---|
| 1 | $3 | n/a | 1 | 2 GB | 40 GB | 2 TB |
| 2 | $6 | $10 | 1 | 4 GB | 80 GB | 4 TB |
| 4 | $12 | $20 | 2 | 8 GB | 160 GB | 8 TB |
| 8 | $24 | $40 | 4 | 16 GB | 320 GB | 16 TB |
| 16 | $48 | $80 | 8 | 32 GB | 640 GB | 32 TB |
| 32 | $96 | $160 | 16 | 64 GB | 1,280 GB | 64 TB |
Three consequences of a linear grid. There is no volume discount, so a 16-slice server costs exactly sixteen times a 1-slice server. There is no renewal cliff, because there is no introductory rate to expire. And cores scale slower than memory: InterServer states that every two slices has access to one CPU core on hosts running between 2.0 and 3.0 GHz, which suits memory-hungry PHP and Node stacks better than CPU-bound work. If your workload is compute-heavy, compare the per-core price against Vultr's High Performance line before committing.
Three US locations and the facilities behind them
Cloud VPS is offered in New York City, Los Angeles and Dallas. InterServer's datacenter page names the buildings, which is rarer than it should be in this industry:
- New York City: the Secaucus, New Jersey campus. InterServer lists its own facility (TEB2, an 8,000 sq ft building) plus space in Equinix NY4 (TEB3), H5 Datacenter (TEB4) and CoreSite NY2 (TEB5). The VPS page does not say which of these a given VPS lands in.
- Los Angeles: Equinix LA3 (LAX1).
- Dallas: 365 Datacenters TX1 (DFW1).
For East Coast users the Secaucus campus is the point: it sits in the same New Jersey corridor as the major exchange datacenters, which is why it shows up in our forex VPS and East Coast VPS pages. Los Angeles and Dallas give the West Coast and the middle of the country a local option. There are no non-US regions for VPS, and no Ashburn, Chicago or Seattle either.
What "managed" means here
InterServer's site makes two statements that read as contradictory until you put them side by side. The VPS FAQ says: "Our VPS are managed, but we manage what we install." Without a control panel, that covers the operating system: boot problems, security updates, patches and reinstalls. With cPanel or DirectAdmin installed, it extends to the panel and the software it installs (web server, FTP, mail). It never covers third-party applications, scripts or forums you add yourself. The VPS page adds a second tier: from 8 slices upward, engineers also help with failed services, databases, uptime and common performance problems.
The practical reading: at 1 to 7 slices you get help when the box will not boot or needs a reinstall, and you own everything above the OS. At 8 slices you get something closer to managed hosting, but still without bundled backups. If you want a provider to own your application layer, look at Cloudways instead.
The four rules you accept at $3
None of these are hidden. They are on InterServer's own VPS page and FAQ. They are also the reason this review does not score the provider higher, and the reason some readers should stop here and go to Vultr.
- No trial, no refund. The FAQ states it plainly: no trial period and no money-back guarantee on VPS hosting. The mitigation is the billing unit: you can cancel from your account and your exposure on a bad first month is $3.
- No hourly billing. Plans are billed month to month. A server you spin up for a two-hour job costs the same as one you run all month. Vultr and DigitalOcean bill by the hour or second; if you build and tear down servers, that is the difference that matters, not the sticker price.
- The 10 Gbps port is shared. Every slice lists a 10 Gbps shared port. Shared means the number is a ceiling for the host, not a guarantee for your server. InterServer does not publish a per-server minimum, and we do not have measurements of our own to fill the gap.
- No independent benchmark data. VPSBenchmarks, the third-party testing site we lean on for other providers, does not list InterServer among its 72 providers as of September 2026. Nothing on this page claims IOPS, CPU scores or latency figures for that reason. If sustained disk or CPU performance is the deciding factor for you, the position is that you will be the one measuring it.
Two smaller gaps sit alongside those rules: InterServer's VPS, FAQ and Windows pages do not mention a snapshot feature or a provider API (the only API on its site is cPanel's own, on the cPanel product page), and backups are a separate paid product rather than a checkbox. What each of those costs depends on who is buying, which is the section below.
What the missing snapshot and API cost, by the size of the buyer
Both absences are planning assumptions rather than surprises: nothing on InterServer's pages promises either feature, and neither is hidden in the fine print. The useful question is what their absence costs, and that answer changes with how many people depend on the machine.
| Buyer | No snapshots | No provider API |
|---|---|---|
| One person, one box | The cost is an evening. A failed change means rebuilding from your own notes and data, which at this size is the same work you did on day one, done faster. Mitigation: keep a database dump and a copy of the configuration somewhere other than that server, and restore one of them once so you know it works. The backup guide builds exactly that job on a $3 InterServer storage slice. | Close to zero. One person administering one server uses SSH and the account portal; an API would be a nicer interface for work you are doing by hand anyway. |
| A small team on one machine | The cost is the working day. A failed change now has a queue behind it: whoever uses the application waits while you rebuild, and the rebuild takes as long as the last rehearsed restore took. This is the size at which a provider that sells snapshots starts being worth the higher plan price, because the difference is one hour against one day. | The cost is the manual step that does not scale. Two people deploying by hand to the same box will eventually overwrite each other or skip a step; the substitute is discipline in the form of a deploy script kept in the repository, which is what a team at this size usually ends up writing anyway. |
| A mid-size business, several environments | Recovery stops being an IT task and becomes a continuity decision, and the reading that follows is that this workload does not belong on a plan without rollback. Either the provider sells snapshots, or the business buys a second machine and a tested procedure and assigns someone to drill it. | This is where the absence decides the purchase. Without a provider API there is no infrastructure as code, no reproducible staging environment, and no way for two engineers to own the same environment without coordinating by hand. If that is the operating model, the hourly providers on the alternatives page are the answer, and their higher plan price is the cost of that model rather than a premium. |
None of that makes the $3 slice a bad product. In a market where every plan page lists twenty features, a server sold without snapshots, without an API and without an hourly meter is a deliberate boundary rather than an oversight: it is built and priced for one always-on box in a US datacenter, and the features it omits are the ones that would raise its price. The question worth answering before you order is not whether it has everything but where you sit in the table above, and whether the row you are in is the row the product was built for.
Who should buy it, and who should not
Buy it if you need a server that runs all month in a US datacenter, you are comfortable administering Linux yourself (or will size to 8 slices for managed support), your traffic fits inside 2 TB per slice, and price predictability matters more than automation. Personal sites, small business sites, WordPress with a caching plugin, a game or Discord bot, a VPN endpoint, and staging servers that stay up all fit. The cheapest VPS page puts this in context against the sub-$5 plans that have less memory.
Skip it if you deploy from code and expect an API, you bill customers by the hour and need the same from your host, you want snapshots before every change, you need a region outside the three listed, or you want a refund window to try before you commit. For the first two, Vultr; for a bigger memory allocation on a slower port, Contabo; for a European server at a Hetzner price, the Hetzner review explains why the cheap plans are EU-only.
If your shortlist is InterServer against one specific competitor, we have InterServer vs Hostwinds and a list of InterServer alternatives for the rest.