CPX52 cloud VPS 12 vCPU, 24 GB RAM
The CPX52 is a Linux cloud VPS with 12 vCPU, 24 GB RAM and 480 GB SSD (Shared AMD EPYC vCPU), from $0.207 per hour with a monthly cap of $129. It deploys in 4 regions (Falkenstein (DE), Helsinki (FI), Nuremberg (DE) and Singapore (SG)) with root SSH access on Ubuntu or Debian, is billed per minute while it exists, and shows its exact price in the configurator before you sign in.
Prices from the live catalog · billed per minute
CPX52 specs and pricing
CPX52 belongs to the CPX line of shared AMD EPYC servers, which offer higher per-core performance than the cost-optimized CX line and are available in more regions.
At $129 per month that works out to $10.76 per vCPU and $5.38 per GB of RAM. Price varies by region: Falkenstein (DE) is the cheapest at $129 per month and Singapore (SG) the most expensive at $172.
CPX52 price by region
Live prices. Servers bill per minute at the hourly rate, up to the monthly cap shown in the configurator.
| Region | Per hour | Monthly cap | |
|---|---|---|---|
| Falkenstein (DE) | $0.207 | $129 | Deploy |
| Helsinki (FI) | $0.207 | $129 | Deploy |
| Nuremberg (DE) | $0.207 | $129 | Deploy |
| Singapore (SG) | $0.275 | $172 | Deploy |
What a CPX52 costs over time
In Falkenstein (DE), the cheapest region for this plan.
| Duration | Cost |
|---|---|
| 1 hour | $0.207 |
| 1 day (24 hours) | $4.97 |
| 1 week (168 hours) | $34.79 |
| Full month | $129 |
What to run on a CPX52
- Large databases and caches
- Analytics and search clusters
- Kubernetes nodes for many pods
- CPU inference of small quantized LLMs
Need AI inference? CPU servers are a poor fit for large language models. Use the pay-per-token inference API for hosted models, or rent a GPU pod billed per minute for custom models and fine-tuning. Single-tenant hardware is available as dedicated bare-metal servers.
CPX52 vs similar plans
Live lowest monthly cap across regions.
| Plan | vCPU | RAM | SSD | CPU type | From / mo |
|---|---|---|---|---|---|
| CPX52 (this page) | 12 | 24 GB | 480 GB | Shared AMD EPYC vCPU | $129 |
| CX43 | 8 | 16 GB | 160 GB | Shared x86 vCPU, cost-optimized | $20.62 |
| CX53 | 16 | 32 GB | 320 GB | Shared x86 vCPU, cost-optimized | $38.44 |
| CAX41 | 16 | 32 GB | 320 GB | Shared Arm64 vCPU, Ampere Altra | $53.02 |
| CPX51 | 16 | 32 GB | 360 GB | Shared AMD EPYC vCPU | $90.82 |
| CPX62 | 16 | 32 GB | 640 GB | Shared AMD EPYC vCPU | $166 |
| CCX33 | 8 | 32 GB | 240 GB | Dedicated AMD EPYC vCPU | $177 |
CPX52 VPS — frequently asked questions
How much does the CPX52 VPS cost?
From $0.207 per hour, capped at $129 per month in Falkenstein (DE). Prices by region range up to $172 per month. Billing is per minute while the server exists, and the configurator shows the hourly and monthly figures before you deploy.
What are the CPX52 specs?
12 vCPU (Shared AMD EPYC vCPU), 24 GB RAM and 480 GB of local SSD storage, with a public IPv4 address and root SSH access.
Where can I deploy a CPX52?
In Falkenstein (DE), Helsinki (FI), Nuremberg (DE) and Singapore (SG). Pick the region closest to your users; prices are listed per region on this page.
What can I run on a CPX52?
Typical workloads: Large databases and caches; Analytics and search clusters; Kubernetes nodes for many pods; CPU inference of small quantized LLMs. Small quantized language models can run on CPU with llama.cpp, but for production LLM traffic a GPU pod or the pay-per-token inference API is far faster and usually cheaper.
Is the CPX52 shared or dedicated vCPU?
Shared. vCPUs share physical cores with other tenants, which keeps the price low and suits bursty workloads such as web apps, CI runners and development boxes. For sustained CPU-bound work choose a dedicated-vCPU CCX plan.
Which operating systems does the CPX52 support?
Ubuntu and Debian, deployed with root SSH access. Install Docker, a web stack or any Linux package after boot.
Deploy a CPX52 in minutes
Pick a region and OS, see the exact hourly and monthly price, and SSH in as root. Billed per minute, no commitment.