Dedicated Hardware Built Around
Your Application
A whole physical server for your web app, API, database or containers. Pick a plan that fits the workload, get full root access, and run it your way. Our engineers are one ticket away when you need them.
Why Modern Applications Outgrow
Generic Virtual Cloud
On a cloud instance your app shares the physical machine with other customers, so a busy neighbour can slow it down. A dedicated server gives every core, every gigabyte of memory and every disk to your application alone.
-
The whole machine is yours: no hypervisor layer and no other tenants on the hardware
-
Local disks: many plans use NVMe drives attached to the server, not storage over the network
-
One monthly price: bandwidth is included in the plan, with no per-gigabyte transfer charges
Dedicated Server vs Cloud Instance
Your code runs on the physical CPU. No other customer shares its cores.
Drives sit inside the server, so there are no network storage limits or burst credits to run out.
Install any supported OS, build your own kernel, load modules and tune it, with full root access.
Each plan lists its bandwidth: unmetered or a monthly allowance on a set port speed, no per-gigabyte fees.
Optimised for the Exact Stacks
You Run in Production
Databases, container clusters or busy APIs: choose the CPU, memory and disks your software needs, then install it your way.
High Throughput Databases
PostgreSQL, MySQL, Redis or MongoDB with enough memory to keep the working set in RAM, on local NVMe or SSD drives.
Microservices & Containers
Docker, Kubernetes or Nomad nodes on servers with up to 64 cores, so many services run side by side.
High Concurrency Web APIs
Go, Node.js, Python or Rust services on a server close to your users, without other tenants competing for the CPU.
Start With What Your Application Needs.
Choose your workload and see three servers from our store that fit it, at their real monthly price. All 26 plans are in the table below.
NVMe storage and at least 32 GB of memory: room for the app, a cache and a small database.
Intel Xeon D-1540
8 cores / 16 threads
- 32 GB DDR3
- 2 x 480 GB NVMe
- Unmetered, 250 Mbps
- 1 usable IPv4
- Canada, United Kingdom, France and more
Intel Core i7-7700K
4 cores / 8 threads
- 32 GB DDR4 ECC
- 2 x 450 GB NVMe
- Unmetered, 300 Mbps
- 1 usable IPv4
- Germany, North America
AMD Ryzen 7 5800X
8 cores / 16 threads
- 64 GB DDR5
- 2 x 512 GB NVMe
- Unmetered, 1 Gbps
- 1 usable IPv4
- Canada, United Kingdom, France and more
128 GB of memory or more on SSD or NVMe, so the busy part of the database stays in RAM.
Dual Intel Xeon E5-2670
- 128 GB DDR4
- 2 x 960 GB SSD
- Unmetered, 1 Gbps
- 5 usable IPv4
- United States
AMD EPYC 7502
32 cores / 64 threads
- 256 GB DDR4
- 2 x 1.92 TB NVMe
- Unmetered, 1 Gbps
- 1 usable IPv4
- United States
AMD Ryzen 9 9950X
16 cores / 32 threads
- 128 GB DDR4
- 2 x 960 GB NVMe
- Unmetered, 1 Gbps
- 1 usable IPv4
- France, Germany, Poland
24 cores or more for Docker or Kubernetes nodes that run many services side by side.
AMD EPYC 7502
32 cores / 64 threads
- 256 GB DDR4
- 2 x 1.92 TB NVMe
- Unmetered, 1 Gbps
- 1 usable IPv4
- United States
AMD EPYC 7402
24 cores / 48 threads
- 128 GB DDR4 ECC
- 1 x 1.92 TB NVMe
- Unmetered, 1 Gbps
- 1 usable IPv4
- United States
Intel Xeon Gold 6312U
24 cores / 48 threads
- 128 GB DDR4
- 2 x 960 GB NVMe
- Unmetered, 1 Gbps
- 1 usable IPv4
- Canada, United Kingdom, France and more
Our lowest-cost servers, for testing, staging, CI runners and internal tools.
Intel Core i3-3220
2 cores / 4 threads
- 16 GB DDR3 ECC
- 1 x 2 TB SATA
- Unmetered, 1 Gbps
- 5 usable IPv4
- United States
Intel Core i7-4790K
4 cores / 8 threads
- 32 GB DDR4
- 4 x 2 TB SATA
- Unmetered, 250 Mbps
- 1 usable IPv4
- United States
Intel Xeon E3-1240v2
- 16 GB DDR3 ECC
- 1 x 480 GB SSD + 1 x 4 TB
- Unmetered, 1 Gbps
- 5 usable IPv4
- United States
All 26 Dedicated Servers in the Store
| Server | Processor | RAM | Storage | Bandwidth | Price / mo | Order |
|---|---|---|---|---|---|---|
|
Core i3-3220
ENTRY LEVEL United States
|
2 cores / 4 threads | 16 GB DDR3 ECC | 1 x 2 TB SATA |
Unmetered 1 Gbps 5 IPv4 |
$74
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Core i7-4790K
ENTRY LEVEL United States
|
4 cores / 8 threads | 32 GB DDR4 | 4 x 2 TB SATA |
Unmetered 250 Mbps 1 IPv4 |
$105
/mo
|
|
|
Server
Location
Included
|
||||||
|
Xeon E3-1240v2
ENTRY LEVEL United States
|
Intel Xeon E3-1240v2 | 16 GB DDR3 ECC | 1 x 480 GB SSD + 1 x 4 TB |
Unmetered 1 Gbps 5 IPv4 |
$120
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Dual Xeon 5520
ENTRY LEVEL United States
|
8 cores / 16 threads | 64 GB DDR3 | 1 x 240 GB SSD + 1 x 4 TB HDD |
Unmetered 1 Gbps 5 IPv4 |
$125
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Xeon E3-1270, 5 IPs
ENTRY LEVEL United States
|
Intel Xeon E3-1270 | 16 GB | 1 TB SATA |
Unmetered 1 Gbps 5 IPv4 |
$133
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Xeon D-1540
ENTRY LEVEL Canada, United Kingdom +2 more
|
8 cores / 16 threads | 32 GB DDR3 | 2 x 480 GB NVMe |
Unmetered 250 Mbps 1 IPv4 |
$138
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Core i7-7700K
BUSINESS Germany, North America
|
4 cores / 8 threads | 32 GB DDR4 ECC | 2 x 450 GB NVMe |
Unmetered 300 Mbps 1 IPv4 |
$160
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Xeon E-2234
BUSINESS United Kingdom, Netherlands
|
4 cores / 8 threads | 16 GB DDR5 | 1 x 240 GB NVMe |
30 TB 1 Gbps 1 IPv4 |
$167
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Ryzen 7 5800X
BUSINESS Canada, United Kingdom +2 more
|
8 cores / 16 threads | 64 GB DDR5 | 2 x 512 GB NVMe |
Unmetered 1 Gbps 1 IPv4 |
$179
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Ryzen 9 9900X
BUSINESS France, Germany, Poland
|
12 cores / 24 threads | 64 GB DDR5 | 2 x 512 GB NVMe |
Unmetered 1 Gbps 1 IPv4 |
$188
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Xeon E-2434
BUSINESS United Kingdom
|
4 cores / 8 threads | 32 GB DDR4 | 2 x 960 GB NVMe |
30 TB 1 Gbps 1 IPv4 |
$220
/mo
|
|
|
Server
Location
Included
|
||||||
|
Dual Xeon E5-2670, 5 IPs
BUSINESS United States
|
Dual Intel Xeon E5-2670 | 128 GB DDR4 | 2 x 960 GB SSD |
Unmetered 1 Gbps 5 IPv4 |
$230
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Xeon E-2334
BUSINESS Japan
|
4 cores / 8 threads | 32 GB DDR4 | 2 x 960 GB NVMe |
30 TB 1 Gbps 1 IPv4 |
$233
/mo
|
|
|
Server
Location
Included
|
||||||
|
EPYC 7502
BUSINESS United States
|
32 cores / 64 threads | 256 GB DDR4 | 2 x 1.92 TB NVMe |
Unmetered 1 Gbps 1 IPv4 |
$239
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Xeon E-2288G
BUSINESS United States
|
8 cores / 16 threads | 32 GB DDR3 | 1 x 480 GB SSD |
20 TB 1 Gbps 1 IPv4 |
$244
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Xeon E-2236
HIGH PERFORMANCE United States
|
6 cores / 12 threads | 64 GB DDR4 | 1 x 480 GB SSD |
20 TB 1 Gbps 1 IPv4 |
$255
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Ryzen 9 9950X
HIGH PERFORMANCE France, Germany, Poland
|
16 cores / 32 threads | 128 GB DDR4 | 2 x 960 GB NVMe |
Unmetered 1 Gbps 1 IPv4 |
$265
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Dual Xeon E5-2670, 1 IP
HIGH PERFORMANCE United States
|
8 cores / 16 threads | 128 GB DDR3 | 2 x 480 GB SSD |
10 TB 1 Gbps 1 IPv4 |
$282
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Dual Xeon E5-2696v3
HIGH PERFORMANCE United States
|
Dual Intel Xeon E5-2696v3 | 128 GB DDR4 | 2 x 2 TB SSD |
Unmetered 1 Gbps 5 IPv4 |
$302
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
EPYC 7402
HIGH PERFORMANCE United States
|
24 cores / 48 threads | 128 GB DDR4 ECC | 1 x 1.92 TB NVMe |
Unmetered 1 Gbps 1 IPv4 |
$321
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Dual Xeon 5630
HIGH PERFORMANCE United States
|
4 cores / 4 threads | 16 GB DDR3 | 4 x 240 GB SSD |
10 TB 1 Gbps 61 IPv4 |
$345
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Xeon Gold 6312U
HIGH PERFORMANCE Canada, United Kingdom +2 more
|
24 cores / 48 threads | 128 GB DDR4 | 2 x 960 GB NVMe |
Unmetered 1 Gbps 1 IPv4 |
$394
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
EPYC Turin 9455
ENTERPRISE France, Germany, Poland
|
48 cores / 96 threads | 128 GB DDR5 | 2 x 1.92 TB SSD |
Unmetered 1 Gbps 1 IPv4 |
$546
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Dual EPYC 7551
ENTERPRISE United States
|
64 cores / 128 threads | 512 GB DDR4 | 1 x 4 TB NVMe |
Unmetered 1 Gbps 5 IPv4 |
$581
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
Xeon E3-1270, 253 IPs
ENTERPRISE United States
|
Intel Xeon E3-1270 | 16 GB DDR3 | 1 x 1 TB SATA |
10 TB 1 Gbps 253 IPv4 |
$586
/mo
|
|
|
Server
Choose a location
Included
|
||||||
|
EPYC 7532
ENTERPRISE Canada, United Kingdom +2 more
|
32 cores / 64 threads | 256 GB DDR4 | 2 x 480 GB SSD |
10 TB 1 Gbps 1 IPv4 |
$800
/mo
|
|
|
Server
Choose a location
Included
|
||||||
| No plans match your search. | ||||||
Run AI on a Server Only You Use.
Host open AI models, connect them to the systems your business already runs, and keep everything they read and write on your own server, in the country you choose.
Host the model
Run open models such as Llama, Mistral, Qwen or Gemma, and give your team a private chat window. Prompts and answers stay on your server and are not shared with a model provider.
- Ollama
- llama.cpp
- vLLM on GPU
- Open WebUI
- Whisper speech to text
Bridge it to your business
MCP servers and APIs let an AI assistant read your CRM, ERP, helpdesk, database, or cPanel and WHM, with the permissions you set. Start read only; the credentials stay on your server, never in the model.
- MCP servers
- REST and GraphQL APIs
- n8n AI workflows
- Webhooks
- Slack, Teams and WhatsApp bots
Store what AI produces
Keep documents converted to Markdown, fields extracted as JSON, embeddings for search and a log of every AI action, next to your own data and backed up on your schedule.
- Markdown and JSON
- PostgreSQL with pgvector
- Qdrant
- Audit logs
- Backups
How much memory does an AI model need?
Approximate memory for open models compressed to 4 bits, the usual way to run them on a server, with the lowest-priced plan that fits.
A CPU server suits one model serving a team. For many users at once, bigger models or image work, choose a GPU dedicated server. Longer prompts need more memory.
Everything We Build With AI and LLMs.
Our engineers plan, install and look after these on your dedicated server, as Hire an Expert work or a custom project. You own the server, the data and the result.
Private AI assistants
- A private chat assistant for your staff
- Assistants that answer from your own documents, with sources
- Department assistants for HR, legal, finance and IT
- Answers and translation in many languages
- Meetings and voice notes turned into text and summaries
Documents and data
- Invoices, receipts and forms read into JSON
- PDFs, Word files and web pages converted to clean Markdown
- Emails, tickets and files sorted and tagged
- Data cleaned, matched and de-duplicated
- Scheduled summaries and reports
Connections, MCP and agents
- MCP servers for your CRM, ERP, helpdesk and databases
- MCP for cPanel and WHM, for resellers and server owners
- AI agents that draft, update and file, with an approval step
- OpenClaw agents set up and secured on your server
- One gateway to hosted models such as OpenAI, Anthropic or Google, with the keys on your server
Automation workflows
- n8n workflows with AI steps
- Email and WhatsApp messages read and routed
- Leads scored and CRM records updated
- Reports posted to Slack or Teams
- Webhooks between your business apps
Search and knowledge
- Search that understands meaning across files, wikis and tickets
- Product search for online stores
- Vector databases with pgvector or Qdrant
- Knowledge bases that stay up to date
- Answers that link back to the source document
Running it well
- Choosing the model and the server size
- Model installs, updates and GPU drivers
- Monitoring, logs and alerts
- Access control, audit trails and backups
- Lower token bills: caching, routing and small local models for simple tasks
- Browser access to servers and desktops with Apache Guacamole
What Businesses Run With AI on Their Own Server.
Common tasks, the parts that run on the server, and what changes for the team. We set each one up to fit your systems.
Accounts: invoices into the books
Invoices arrive by email. A model reads each one, writes supplier, date, lines and totals as JSON, and your accounting software imports it. Staff only check the exceptions.
On the server: Email inbox, OCR, an open model, n8n, your accounting API
Support: replies drafted from your answers
Each new ticket gets a suggested reply built from your help articles and past answers, with the sources listed. An agent edits and sends it.
On the server: Helpdesk via MCP or API, a vector database, an open model
Sales: ask the CRM in plain words
"Which deals over $10,000 are due to close this month?" A read-only MCP server lets an AI assistant answer from live CRM data without anyone building a report.
On the server: MCP server on your CRM, an AI assistant, audit log
Operations: manuals that answer back
Technicians ask how to fix a machine and get the steps from your manuals and procedures, with the page they came from.
On the server: Manuals converted to Markdown, pgvector or Qdrant, an open model
Regulated teams: private by design
Legal, healthcare and finance teams use an assistant on data that may not go to an outside AI service. The server sits in the country you choose.
On the server: Open model, private chat, access control, your choice of location
Online stores: search that understands
Shoppers type "waterproof jacket for trekking under $80" and find it. New product descriptions are drafted from the specs for a person to approve.
On the server: Product data as JSON, embeddings, an open model, your store API
Hosting resellers: manage accounts by chat
Create accounts, reset passwords and check usage by asking an AI assistant, through an MCP server on the cPanel and WHM API, limited to your own accounts.
On the server: MCP on cPanel and WHM, your AI assistant, audit log
Management: the week in one page
Every Monday, tickets, sales and server alerts are summarised into one short report and posted to Teams or Slack.
On the server: n8n schedule, an open model, your apps' APIs
Planning a bigger project? See custom business solutions.
Engineered for Your
Production Stack
You own the software stack, from the operating system and kernel settings to the RAID layout. Here is what every server includes and what you can add.
See the plansPhysical Resources, Not Shared
Every core, every gigabyte of memory and every drive in the server is for your application. No hypervisor and no other customers on the machine.
Full Root, IPMI on Request
Full root access to the operating system you choose at order. IPMI or KVM console access is usually free on request, where the data centre offers it.
NVMe and SSD Plans
Many plans use local NVMe or SSD drives for databases and busy file systems; each plan lists its exact disks. Set up RAID 1 or RAID 10 on plans with two or more drives.
Free IPv6, Extra IPv4
Every plan has IPv6-ready routing, and IPv6 addresses are free on request. Each plan lists its usable IPv4 addresses; extra IPv4 subnets from /29 to /24 can be ordered.
DDoS Protection Where Offered
DDoS protection is available on selected servers, mostly in the UK and Europe. Ask us before you order and we will confirm it for the server you choose.
AI, API and Automation Help
Our engineers can set up AI models, n8n workflows, APIs and MCP servers for AI assistants on your server, as paid Hire an Expert work.
Who Builds on a Dedicated Application Server
Teams who need steady performance, full control and their data on their own hardware.
High-Growth SaaS Platforms
Multi-tenant web applications that need steady response times and a monthly bill that does not grow with every gigabyte of traffic.
Enterprise Databases & Analytics
PostgreSQL, ClickHouse and Elasticsearch, which run best with plenty of memory and local NVMe or SSD drives.
High-Volume eCommerce
Magento, WooCommerce or headless stores with large catalogues, busy search and seasonal traffic peaks.
CI/CD & Compute Workers
Build and test pipelines, media transcoding queues and AI inference on CPU, which keep every core busy and need fast local disks.
From Order to Running Application.
Four steps, who does each one, and how long it takes.
-
Step 1
Choose a server
Use the workload picks or the full table. Not sure? Open a ticket with your stack and expected traffic and we suggest a plan.
-
Step 2
Order and verify
Pick the operating system and pay. New accounts may be asked for an ID check before the server is set up.
-
Step 3
We deliver
The server is set up with your operating system and the root login is sent to you. Ask for free IPv6 or IPMI access at any time.
-
Step 4
Deploy and grow
Install your application. Free Basic Managed Support covers quick tasks; migrations, AI setups and tuning are Hire an Expert work.
Moving Your Production Stack Is
Routine Engineering
Our engineers can move your application for you: copy the files and databases, set up the services, test everything on the new server, then switch DNS in your quietest hours. On these self-managed servers a migration is Hire an Expert work, quoted before we start.
Ask for a migration quoteReal Feedback From the Businesses We Support
See how our engineering support and infrastructure keep real businesses online.
Dedicated server, with engineers on call
Self-managed means you decide how the server is built. It does not mean you are on your own. Every dedicated server includes Free Basic Managed Support, and when a job needs more hands you can bring our engineers in for 15 minutes, one task or a full month, then stop again.
Free Basic Managed Support
Complimentary, best-effort help with quick tasks of about 5 minutes: boot and access checks, OS reinstalls, IPMI/KVM help, IP and DNS pointing, service restarts and first-level troubleshooting.
What is included →Plug-and-play expert help
Hardening, panel installs, migrations, RAID or firewall work. From $8 per 15 minutes or $28 per hour, with no contract and no monthly commitment.
See on-demand support →AI, automation & custom builds
Turn the box into a product: private AI models, n8n automations, CI pipelines, game or streaming stacks built and handed over to you.
Plan a project →Frequently Asked Questions
Quick answers about the plans, support and billing.
Run Your Application on a Server of Its Own
Pick a plan from $74 a month, or tell our engineers about your stack and we suggest one.