FSDOM FSDOM
  • Support
    • FAQ
    • Knowledgebase
    • Guides
    • Contact Us
    • Live Chat
    • Billing
    • -----
    • Server status
    • Looking glass
  • English English
    english English
    german Deutsch
    spanish Español
    french Français
    portuguese-br Português
    italian Italiano
    dutch Nederlands
    swedish Svenska
    romanian Română
    czech Čeština
    hungarian Magyar
    turkish Türkçe
    ukrainian Українська
    arabic العربية
    polish Polski
    portuguese Português
    No Records Found
  • USD
    USD ($)
    EUR (€)
    GBP (£)
    RON (lei)
    PLN (zł)
    SEK (kr)
FSDOM
  • VPS hosting
    Virtual servers
    • Cloud-Check--Streamline-Core (8).svg VPSInstant, production-ready KVM VPS
    • Ai-Chip-Spark--Streamline-Core (2).svg AMD Ryzen VDS4.5 GHz Ryzen with dedicated CPU
    • Database-Check--Streamline-Core (4).svg Storage VPSHigh-capacity storage servers.
    • Insert-Cloud-Video--Streamline-Core (11).svg GPU VPSAffordable industry-leading GPU.
    • nodejs_logo_icon_214669.svg Node.js VPS hostingLaunch apps with ease.
    • Web--Streamline-Core (5).svg Web hosting VPSHosting optimized for websites
    Application hosting
    • OpenClawAI agent hosting
    • Docker hostingContainer ready VPS
    • Self-hosted n8nWorkflow automation
    • Forex VPSLow latency trading
    • Hermes AgentPersonal 24/7 AI agent
    • Coolify hostingSelf-hosted PaaS
    • Claude CodeNEWRun AI coding
    • ispmanager panelVPS with panel included
    • Windows RDP VPSRemote desktop VPS
  • GPU servers
  • Dedicated servers
  • Hermes VPS
  • Explore
    Company
    • Our storyBuilt on trust.
    • ContactWe’re here to help.
    • Guides Build with confidence.
    • Blog & newsLet's stay connected!
    • Affiliate programTurn traffic into revenue.
    • Reseller programSell more, support less.
    Features
    • Database-Server-2--Streamline-Core (4).svg Data center infrastructureSecure and trusted datacenters with enterprise hardware.
    • Ai-Generate-Variation-Spark--Streamline-Core (4).svg One-click apps and operating systemsExplore our library of OS templates and applications
  • Login
Accelerate the future

Dedicated GPU VPS hosting

Instantly deploy production-ready NVIDIA Blackwell servers, built for AI training and real-time inference.

  • Dedicated performance with full GPU passthrough
  • Unrivaled speed with AMD EPYC & NVMe storage
See plans
Starting at $129.00 /month
Our customers say Excellent
reviews
Costs, kept simple

Rent GPU servers with fixed, affordable pricing

Ready in seconds 7-day money-back guaranteeA risk-free way to try Funom. Covers the GPU VPS plan on your first order. Cancel anytime
  • All
  • 16GB vRAM
  • 20 - 24GB vRAM
  • >32GB vRAM
Billing Cycle

GPU.T4

$149.00 Save  13 %
$129.00 Monthly
  • Dedicated GPU
  • Tesla T4

  • 16 GB GDDR6vRAM
  • 2560CUDA CORES
  • Virtual Server
  • 8 vCPUAMD EPYC
  • 32 GBECC MEMORY
  • 250 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Order Now Unavailable
1 Available

GPU.PRO2000

Featured
$209.00 Save  19 %
$169.00 Monthly
  • Dedicated GPU
  • RTX PRO 2000 Blackwell

  • 16 GB GDDR7 ECCvRAM
  • 4352CUDA CORES
  • Virtual Server
  • 12 vCPUAMD EPYC
  • 48 GBECC MEMORY
  • 300 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Order Now Unavailable
1 Available

GPU.ADA4000SFF

$249.00 Save  20 %
$199.00 Monthly
  • Dedicated GPU
  • RTX 4000 SFF Ada

  • 20 GB GDDR6 ECCvRAM
  • 6144CUDA CORES
  • Virtual Server
  • 16 vCPUAMD EPYC
  • 64 GBECC MEMORY
  • 350 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Sold Out Unavailable
0 Available

GPU.PRO4000SFF

$299.00 Save  20 %
$239.00 Monthly
  • Dedicated GPU
  • RTX PRO 4000 Blackwell

  • 24 GB GDDR7 ECCvRAM
  • 8960CUDA CORES
  • Virtual Server
  • 16 vCPUAMD EPYC
  • 64 GBECC MEMORY
  • 400 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Sold Out Unavailable
0 Available

GPU.PRO4500

$499.00 Save  20 %
$399.00 Monthly
  • Dedicated GPU
  • RTX PRO 4500 Blackwell

  • 32 GB GDDR7 ECCvRAM
  • 10496CUDA CORES
  • Virtual Server
  • 16 vCPUAMD EPYC
  • 64 GBECC MEMORY
  • 450 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Sold Out Unavailable
0 Available

GPU.PRO5000

$699.00 Save  20 %
$559.00 Monthly
  • Dedicated GPU
  • RTX PRO 5000 Blackwell

  • 48 GB GDDR7 ECCvRAM
  • 14080CUDA CORES
  • Virtual Server
  • 32 vCPUAMD EPYC
  • 96 GBECC MEMORY
  • 500 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Sold Out Unavailable
0 Available

GPU.PRO6000

$1,199.00 Save  19 %
$969.00 Monthly
  • Dedicated GPU
  • RTX PRO 6000 Blackwell

  • 96 GB GDDR7 ECCvRAM
  • 24064CUDA CORES
  • Virtual Server
  • 32 vCPUAMD EPYC
  • 128 GBECC MEMORY
  • 650 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Sold Out Unavailable
0 Available

GPU.T4

$149.00 Save  13 %
$129.00 Monthly
  • Dedicated GPU
  • Tesla T4

  • 16 GB GDDR6vRAM
  • 2560CUDA CORES
  • Virtual Server
  • 8 vCPUAMD EPYC
  • 32 GBECC MEMORY
  • 250 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Order Now Unavailable
1 Available

GPU.PRO2000

Featured
$209.00 Save  19 %
$169.00 Monthly
  • Dedicated GPU
  • RTX PRO 2000 Blackwell

  • 16 GB GDDR7 ECCvRAM
  • 4352CUDA CORES
  • Virtual Server
  • 12 vCPUAMD EPYC
  • 48 GBECC MEMORY
  • 300 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Order Now Unavailable
1 Available

GPU.ADA4000SFF

$249.00 Save  20 %
$199.00 Monthly
  • Dedicated GPU
  • RTX 4000 SFF Ada

  • 20 GB GDDR6 ECCvRAM
  • 6144CUDA CORES
  • Virtual Server
  • 16 vCPUAMD EPYC
  • 64 GBECC MEMORY
  • 350 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Sold Out Unavailable
0 Available

GPU.PRO4000SFF

$299.00 Save  20 %
$239.00 Monthly
  • Dedicated GPU
  • RTX PRO 4000 Blackwell

  • 24 GB GDDR7 ECCvRAM
  • 8960CUDA CORES
  • Virtual Server
  • 16 vCPUAMD EPYC
  • 64 GBECC MEMORY
  • 400 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Sold Out Unavailable
0 Available

GPU.PRO4500

$499.00 Save  20 %
$399.00 Monthly
  • Dedicated GPU
  • RTX PRO 4500 Blackwell

  • 32 GB GDDR7 ECCvRAM
  • 10496CUDA CORES
  • Virtual Server
  • 16 vCPUAMD EPYC
  • 64 GBECC MEMORY
  • 450 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Sold Out Unavailable
0 Available

GPU.PRO5000

$699.00 Save  20 %
$559.00 Monthly
  • Dedicated GPU
  • RTX PRO 5000 Blackwell

  • 48 GB GDDR7 ECCvRAM
  • 14080CUDA CORES
  • Virtual Server
  • 32 vCPUAMD EPYC
  • 96 GBECC MEMORY
  • 500 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Sold Out Unavailable
0 Available

GPU.PRO6000

$1,199.00 Save  19 %
$969.00 Monthly
  • Dedicated GPU
  • RTX PRO 6000 Blackwell

  • 96 GB GDDR7 ECCvRAM
  • 24064CUDA CORES
  • Virtual Server
  • 32 vCPUAMD EPYC
  • 128 GBECC MEMORY
  • 650 GB NVMeSTORAGE
  • Unmetered bandwidth
  • IPv4 & IPv6IPv6 is currently unavailable in France, Finland or the Netherlands. included
Sold Out Unavailable
0 Available

*VAT excluded.

INCLUDED WITH EVERY PLAN

✓ No setup fees ✓1 Gbps network
✓ Free server monitoring ✓ Firewall management ✓24/7 support ✓ KVM virtualization

GPU products are in high demand at the moment. Fill the form to get notified as soon as your preferred GPU model is back in stock.

Trusted by 17k developers & businesses in 97+ countries

Dedicated GPU VPS hosting - Trusted by 17k developers & businesses in 97+ countries

Dedicated GPU VPS hosting at every scale

  • RTX PRO 2000 Blackwell
  • Tesla T4
  • RTX 4000 ADA SFF
  • RTX PRO 4000 Blackwell
  • RTX PRO 4500 Blackwell
  • RTX PRO 5000 Blackwell
  • RTX PRO 6000 Blackwell
RTX PRO 2000 Blackwell

The Blackwell entry point and the natural pick for a first production deployment. FP4 precision squeezes remarkable throughput from a compact card, so chat assistants and RAG services respond quickly without the cost of hardware you don't need yet. Start here and grow into the lineup.

  • Interactive inference on 7B to 14B models
  • AV1 encode for streaming pipelines
Explore more
Dedicated GPU VPS hosting - Dedicated GPU VPS hosting at every scale
GPU memory16 GB GDDR7 ECC
CUDA cores4,352
Tensor cores136 (5th gen)
AI performance 545 TOPS
FP32 performance 17 TFLOPS
Memory bandwidth 288 GB/s
Tesla T4

A proven workhorse that's been earning its place in data centers for years. The T4 remains the efficient choice for steady, well-defined jobs: scheduled transcoding, embedding generation and classification models that run around a fixed budget. Reliable silicon for work that values consistency over peak speed.

  • INT8 and FP16 inference on small models
  • NVENC hardware video transcoding
Explore more
Dedicated GPU VPS hosting - Dedicated GPU VPS hosting at every scale
GPU memory16 GB GDDR6
CUDA cores2,560
Tensor cores320 (2nd gen)
AI performance 130 TOPS
FP32 performance 8.1 TFLOPS
Memory bandwidth 320 GB/s
RTX 4000 ADA SFF

The image generation specialist. Twenty gigabytes of vRAM means Stable Diffusion and ComfyUI run at full resolution with several checkpoints loaded, and Ada's dual encoders make it just as capable in a video pipeline. Creators and render services get the most from this card.

  • High-resolution diffusion with room for LoRAs
  • AV1 encode on dual NVENC engines
Explore more
Dedicated GPU VPS hosting - Dedicated GPU VPS hosting at every scale
GPU memory20 GB GDDR6 ECC
CUDA cores6,144
Tensor cores192 (4th gen)
AI performance 306 TOPS
FP32 performance 19.2 TFLOPS
Memory bandwidth 280 GB/s
RTX PRO 4000 Blackwell

Blackwell throughput in a compact, efficient package. This is the sweet spot for teams running 30B-class models quantized, where GDDR7 bandwidth keeps token generation brisk and 24 GB leaves working room for context. A serious amount of capability per watt and per dollar.

  • Runs 30B-class models quantized
  • LoRA fine-tuning on mid-size models
Explore more
Dedicated GPU VPS hosting - Dedicated GPU VPS hosting at every scale
GPU memory24 GB GDDR7 ECC
CUDA cores8,960
Tensor cores280 (5th gen)
AI performance 770 TOPS
FP32 performance 24 TFLOPS
Memory bandwidth 432 GB/s
RTX PRO 4500 Blackwell

Built for inference that serves real users all day. Independent benchmarks put 14B models around 75 tokens per second here, comfortably ahead of reading speed, and 32 GB holds mid-size models with generous context. The dependable middle of the lineup for production chat and vision AI.

  • Real-time chat under multi-user load
  • Video analytics and vision pipelines
Explore more
Dedicated GPU VPS hosting - Dedicated GPU VPS hosting at every scale
GPU memory32 GB GDDR7 ECC
CUDA cores10,496
Tensor cores328 (5th gen)
AI performance 1,600 TOPS
FP32 performance 51 TFLOPS
Memory bandwidth 800 GB/s
RTX PRO 5000 Blackwell

The card for teams whose models and traffic are both growing. With 48 GB of GDDR7, 70B-class models fit in 4-bit quantization, full fine-tuning opens up on mid-size models and the bandwidth keeps queues moving through long contexts. Production capacity with headroom to spare.

  • 70B-class models in 4-bit quantization
  • Sustained throughput for busy inference queues
Explore more
Dedicated GPU VPS hosting - Dedicated GPU VPS hosting at every scale
GPU memory48 GB GDDR7 ECC
CUDA cores14,080
Tensor cores440 (5th gen)
AI performance 2,142 TOPS
FP32 performance 72 TFLOPS
Memory bandwidth 1,344 GB/s
RTX PRO 6000 Blackwell

The flagship, and a category of its own. Ninety-six gigabytes on a single card runs 70B models without sharding, at higher precision than smaller cards allow, with capacity left for long contexts and concurrent sessions. For the largest open models, this is the whole answer on one slot.

  • 70B models served without sharding
  • Fine-tuning at scale on one card
Explore more
Dedicated GPU VPS hosting - Dedicated GPU VPS hosting at every scale
GPU memory96 GB GDDR7 ECC
CUDA cores24,064
Tensor cores752 (5th gen)
AI performance 4,000 TOPS
FP32 performance 120 TFLOPS
Memory bandwidth 1,597 GB/s
Dedicated GPU VPS hosting - No setup fees
No setup fees
Dedicated GPU VPS hosting - Full root access
Full root access
Dedicated GPU VPS hosting - 1 Gbps network
1 Gbps network
Dedicated GPU VPS hosting - Scale up anytime
Scale up anytime
Dedicated GPU VPS hosting - Unmetered traffic
Unmetered traffic
Ready for what's next

Host AI and media pipelines

  • Self-hosted LLM inference
  • Model fine-tuning and training
  • Stable Diffusion and image generation
  • AI agents with local models
  • 3D rendering
  • Video encoding and streaming
Self-hosted LLM inference
Serve Llama or DeepSeek through vLLM or Ollama and keep every prompt on hardware you control. Quantized models run fast on FP4-capable tensor cores.
Model fine-tuning and training
Fine-tune open AI models on your own data with PyTorch or TensorFlow. The card is yours for the entire run, so long training jobs never compete with anyone else's schedule.
Stable Diffusion and image generation
Generate at full resolution with Stable Diffusion or ComfyUI and batch large jobs overnight. Dedicated vRAM means output quality never drops to fit a shared slice.
AI agents with local models
Pair OpenClaw or Hermes Agent with a local model through Ollama and your agent reasons privately, with no per-token bill attached.
3D rendering
Move Blender Cycles or V-Ray work onto RT cores built for it. Render farms of one are entirely realistic when the card isn't shared.
Video encoding and streaming
Run FFmpeg on hardware NVENC with AV1 support and keep the CPU free for the application itself. Transcoding pipelines and streaming platforms both fit here.
Launch your GPU
Dedicated GPU VPS hosting - Self-hosted LLM inference
Secure global datacenters

Deploy servers across Europe, UK & USA

Dedicated GPU VPS hosting - UK

UK

London

Dedicated GPU VPS hosting - germany

Germany

Frankfurt
Dedicated GPU VPS hosting - united_states

United States

New York
Dedicated GPU VPS hosting - netherlands

Netherlands

Amsterdam
Dedicated GPU VPS hosting - france

France

Paris
Dedicated GPU VPS hosting - spain

Spain

Madrid
Dedicated GPU VPS hosting - Finland

Finland

Helsinki
Dedicated GPU VPS hosting - poland

Poland

Warsaw
Dedicated GPU VPS hosting - romania

Romania

Bucharest
Dedicated GPU VPS hosting - italy

Italy

Milan
Dedicated GPU VPS hosting - UK
UK
 -

London

Dedicated GPU VPS hosting - germany
Germany
 - Frankfurt
Dedicated GPU VPS hosting - united_states
United States
 - New York
Dedicated GPU VPS hosting - netherlands
Netherlands
 - Amsterdam
Dedicated GPU VPS hosting - france
France
 - Paris
Dedicated GPU VPS hosting - spain
Spain
 - Madrid
Dedicated GPU VPS hosting - Finland
Finland
 - Helsinki
Dedicated GPU VPS hosting - poland
Poland
 - Warsaw
Dedicated GPU VPS hosting - romania
Romania
 - Bucharest
Dedicated GPU VPS hosting - italy
Italy
 - Milan

Why teams choose a cloud GPU server

Undivided GPU speed

The card is attached to one server, so throughput matches the datasheet on every run.

A server, not a session

Models and environments stay on disk between runs, ready the moment you reconnect.

Compute sized for GPU

EPYC cores and NVMe storage keep data moving fast enough to keep the card busy.

Your data stays here

Model weights and prompts stay on the server you rent, under your own access controls.

Operating systems and applications

One-click templates, ready to go

Your server comes online already configured. Choose what you need from our one-click library and start working instead of spending an afternoon on setup.
Dedicated GPU VPS hosting - NodeJS
NodeJS
Dedicated GPU VPS hosting - Coolify
Coolify
Dedicated GPU VPS hosting - OpenClaw
OpenClaw
Dedicated GPU VPS hosting - n8n
n8n
Dedicated GPU VPS hosting - Docker
Docker
Dedicated GPU VPS hosting - listmonk
listmonk
Dedicated GPU VPS hosting - Hermes Agent
Hermes Agent
Dedicated GPU VPS hosting - Wordpress
Wordpress
Dedicated GPU VPS hosting - Portainer
Portainer
Dedicated GPU VPS hosting - Arcane
Arcane
Dedicated GPU VPS hosting - cPanel
cPanel
Dedicated GPU VPS hosting - Easypanel
Easypanel
Dedicated GPU VPS hosting - ispmanager
ispmanager
Dedicated GPU VPS hosting - Dokploy
Dokploy
Dedicated GPU VPS hosting - Hestia
Hestia
Dedicated GPU VPS hosting - CyberPanel
CyberPanel
Dedicated GPU VPS hosting - ZeroClaw
ZeroClaw
Dedicated GPU VPS hosting - Kubernetes
Kubernetes
Dedicated GPU VPS hosting - Wireguard
Wireguard
Dedicated GPU VPS hosting - Gemini CLI
Gemini CLI
Dedicated GPU VPS hosting - Claude Code
Claude Code
Dedicated GPU VPS hosting - Grafana
Grafana
Dedicated GPU VPS hosting - Pangolin
Pangolin
Dedicated GPU VPS hosting - Magento
Magento
Dedicated GPU VPS hosting - aaPanel
aaPanel
Dedicated GPU VPS hosting - Webmin
Webmin
Dedicated GPU VPS hosting - Miniconda
Miniconda
Dedicated GPU VPS hosting - OpenWebUI
OpenWebUI
Dedicated GPU VPS hosting - Plesk
Plesk
Dedicated GPU VPS hosting - VSFTPD
VSFTPD
Dedicated GPU VPS hosting - Pritunl
Pritunl
Dedicated GPU VPS hosting - odoo
odoo
Dedicated GPU VPS hosting - Uptime Kuma
Uptime Kuma
Dedicated GPU VPS hosting - Librechat
Librechat
Dedicated GPU VPS hosting - AzuraCast
AzuraCast
Dedicated GPU VPS hosting - Clustercontrol
Clustercontrol
Dedicated GPU VPS hosting - Filecloud
Filecloud
Dedicated GPU VPS hosting - Immich
Immich
Dedicated GPU VPS hosting - Microsoft SQL
Microsoft SQL
Dedicated GPU VPS hosting - OpenCode
OpenCode
Dedicated GPU VPS hosting - Crowdsec
Crowdsec
Dedicated GPU VPS hosting - Prometheus
Prometheus
Dedicated GPU VPS hosting - Jellyfin
Jellyfin
Dedicated GPU VPS hosting - Nextcloud
Nextcloud
Dedicated GPU VPS hosting - NodeJS
NodeJS
Dedicated GPU VPS hosting - Coolify
Coolify
Dedicated GPU VPS hosting - OpenClaw
OpenClaw
Dedicated GPU VPS hosting - n8n
n8n
Dedicated GPU VPS hosting - Docker
Docker
Dedicated GPU VPS hosting - listmonk
listmonk
Dedicated GPU VPS hosting - Hermes Agent
Hermes Agent
Dedicated GPU VPS hosting - Wordpress
Wordpress
Dedicated GPU VPS hosting - Portainer
Portainer
Dedicated GPU VPS hosting - Arcane
Arcane
Dedicated GPU VPS hosting - cPanel
cPanel
Dedicated GPU VPS hosting - Easypanel
Easypanel
Dedicated GPU VPS hosting - ispmanager
ispmanager
Dedicated GPU VPS hosting - Dokploy
Dokploy
Dedicated GPU VPS hosting - Hestia
Hestia
Dedicated GPU VPS hosting - CyberPanel
CyberPanel
Dedicated GPU VPS hosting - ZeroClaw
ZeroClaw
Dedicated GPU VPS hosting - Kubernetes
Kubernetes
Dedicated GPU VPS hosting - Wireguard
Wireguard
Dedicated GPU VPS hosting - Gemini CLI
Gemini CLI
Dedicated GPU VPS hosting - Claude Code
Claude Code
Dedicated GPU VPS hosting - Grafana
Grafana
Dedicated GPU VPS hosting - Pangolin
Pangolin
Dedicated GPU VPS hosting - Magento
Magento
Dedicated GPU VPS hosting - aaPanel
aaPanel
Dedicated GPU VPS hosting - Webmin
Webmin
Dedicated GPU VPS hosting - Miniconda
Miniconda
Dedicated GPU VPS hosting - OpenWebUI
OpenWebUI
Dedicated GPU VPS hosting - Plesk
Plesk
Dedicated GPU VPS hosting - VSFTPD
VSFTPD
Dedicated GPU VPS hosting - Pritunl
Pritunl
Dedicated GPU VPS hosting - odoo
odoo
Dedicated GPU VPS hosting - Uptime Kuma
Uptime Kuma
Dedicated GPU VPS hosting - Librechat
Librechat
Dedicated GPU VPS hosting - AzuraCast
AzuraCast
Dedicated GPU VPS hosting - Clustercontrol
Clustercontrol
Dedicated GPU VPS hosting - Filecloud
Filecloud
Dedicated GPU VPS hosting - Immich
Immich
Dedicated GPU VPS hosting - Microsoft SQL
Microsoft SQL
Dedicated GPU VPS hosting - OpenCode
OpenCode
Dedicated GPU VPS hosting - Crowdsec
Crowdsec
Dedicated GPU VPS hosting - Prometheus
Prometheus
Dedicated GPU VPS hosting - Jellyfin
Jellyfin
Dedicated GPU VPS hosting - Nextcloud
Nextcloud
Dedicated GPU VPS hosting - NodeJS
NodeJS
Dedicated GPU VPS hosting - Coolify
Coolify
Dedicated GPU VPS hosting - OpenClaw
OpenClaw
Dedicated GPU VPS hosting - n8n
n8n
Dedicated GPU VPS hosting - Docker
Docker
Dedicated GPU VPS hosting - listmonk
listmonk
Dedicated GPU VPS hosting - Hermes Agent
Hermes Agent
Dedicated GPU VPS hosting - Wordpress
Wordpress
Dedicated GPU VPS hosting - Portainer
Portainer
Dedicated GPU VPS hosting - Arcane
Arcane
Dedicated GPU VPS hosting - cPanel
cPanel
Dedicated GPU VPS hosting - Easypanel
Easypanel
Dedicated GPU VPS hosting - ispmanager
ispmanager
Dedicated GPU VPS hosting - Dokploy
Dokploy
Dedicated GPU VPS hosting - Hestia
Hestia
Dedicated GPU VPS hosting - CyberPanel
CyberPanel
Dedicated GPU VPS hosting - ZeroClaw
ZeroClaw
Dedicated GPU VPS hosting - Kubernetes
Kubernetes
Dedicated GPU VPS hosting - Wireguard
Wireguard
Dedicated GPU VPS hosting - Gemini CLI
Gemini CLI
Dedicated GPU VPS hosting - Claude Code
Claude Code
Dedicated GPU VPS hosting - Grafana
Grafana
Dedicated GPU VPS hosting - Pangolin
Pangolin
Dedicated GPU VPS hosting - Magento
Magento
Dedicated GPU VPS hosting - aaPanel
aaPanel
Dedicated GPU VPS hosting - Webmin
Webmin
Dedicated GPU VPS hosting - Miniconda
Miniconda
Dedicated GPU VPS hosting - OpenWebUI
OpenWebUI
Dedicated GPU VPS hosting - Plesk
Plesk
Dedicated GPU VPS hosting - VSFTPD
VSFTPD
Dedicated GPU VPS hosting - Pritunl
Pritunl
Dedicated GPU VPS hosting - odoo
odoo
Dedicated GPU VPS hosting - Uptime Kuma
Uptime Kuma
Dedicated GPU VPS hosting - Librechat
Librechat
Dedicated GPU VPS hosting - AzuraCast
AzuraCast
Dedicated GPU VPS hosting - Clustercontrol
Clustercontrol
Dedicated GPU VPS hosting - Filecloud
Filecloud
Dedicated GPU VPS hosting - Immich
Immich
Dedicated GPU VPS hosting - Microsoft SQL
Microsoft SQL
Dedicated GPU VPS hosting - OpenCode
OpenCode
Dedicated GPU VPS hosting - Crowdsec
Crowdsec
Dedicated GPU VPS hosting - Prometheus
Prometheus
Dedicated GPU VPS hosting - Jellyfin
Jellyfin
Dedicated GPU VPS hosting - Nextcloud
Nextcloud
Dedicated GPU VPS hosting - CentOS
CentOS
Dedicated GPU VPS hosting - AlmaLinux
AlmaLinux
Dedicated GPU VPS hosting - Debian
Debian
Dedicated GPU VPS hosting - Ubuntu
Ubuntu
Dedicated GPU VPS hosting - Archlinux
Archlinux
Dedicated GPU VPS hosting - openSUSE
openSUSE
Dedicated GPU VPS hosting - Rocky Linux
Rocky Linux
Dedicated GPU VPS hosting - Fedora
Fedora
Dedicated GPU VPS hosting - Alpine Linux
Alpine Linux
Dedicated GPU VPS hosting - Windows
Windows
Dedicated GPU VPS hosting - FreeBSD
FreeBSD
Dedicated GPU VPS hosting - CentOS
CentOS
Dedicated GPU VPS hosting - AlmaLinux
AlmaLinux
Dedicated GPU VPS hosting - Debian
Debian
Dedicated GPU VPS hosting - Ubuntu
Ubuntu
Dedicated GPU VPS hosting - Archlinux
Archlinux
Dedicated GPU VPS hosting - openSUSE
openSUSE
Dedicated GPU VPS hosting - Rocky Linux
Rocky Linux
Dedicated GPU VPS hosting - Fedora
Fedora
Dedicated GPU VPS hosting - Alpine Linux
Alpine Linux
Dedicated GPU VPS hosting - Windows
Windows
Dedicated GPU VPS hosting - FreeBSD
FreeBSD
Dedicated GPU VPS hosting - CentOS
CentOS
Dedicated GPU VPS hosting - AlmaLinux
AlmaLinux
Dedicated GPU VPS hosting - Debian
Debian
Dedicated GPU VPS hosting - Ubuntu
Ubuntu
Dedicated GPU VPS hosting - Archlinux
Archlinux
Dedicated GPU VPS hosting - openSUSE
openSUSE
Dedicated GPU VPS hosting - Rocky Linux
Rocky Linux
Dedicated GPU VPS hosting - Fedora
Fedora
Dedicated GPU VPS hosting - Alpine Linux
Alpine Linux
Dedicated GPU VPS hosting - Windows
Windows
Dedicated GPU VPS hosting - FreeBSD
FreeBSD
Premium features, as standard

Included with every GPU VPS

Everything a production server needs is ready from the first boot.
Unmetered bandwidth
GPU passthrough
100% NVMe storage
Full driver control
Public IPv4 + IPv6
Web terminal
KVM + full root
ECC memory
Firewall control
7-day refund
DDoS protection
24/7 support
More than compute

The human side of high performance

When your work depends on the hardware, the people behind it matter. Our certified specialists support you 24/7 and in multiple languages, backed by 17+ years of running enterprise infrastructure at scale.

Live chat
Help-center
Guides
Open ticket
Dedicated GPU VPS hosting - The human side of high performance
Backed by trust. Measured by results.

Your success story begins here

Reviews rating

Good price for what you get

2025-12-29

"I´ve been using their ops for a while and it´s been great. Performance is good and dependable, and the price makes it even better. everything just works smoothly, and they are very responsive if you need anything."
5 stars

Dedicated GPU VPS hosting - Good price for what you get ”
Alejandro Morales
Hostadvice

Setup was quick and the...

2025-09-01

"Setup was quick and the VPS has been running smoothly since. Customer support answered my questions without making things complicated, which I honestly liked."

5 stars
Dedicated GPU VPS hosting - Setup was quick and the... ”
Paul Eisenberg
Facebook

Amazing team and company

2025-03-01

"Amazing team and company, as they continue to grow the availability of the services to other regions of the world they continue to expand on excellence and efficiency."

5 stars
Dedicated GPU VPS hosting - Amazing team and company ”
Vicente Muñoz
Google

Good quality

2025-03-09

"Fast VPS, no complaints.."

5 stars
Dedicated GPU VPS hosting - Good quality ”
Jona Kaiser
Facebook

Good service for web hosting

2025-06-24

"Good service for web hosting. Easy setup and fair price."

5 stars
Dedicated GPU VPS hosting - Good service for web hosting ”
Christine
Google

Excellent provider

2026-01-25

"I’ve been with funom for almost a year now, and their servers are great. I have a few Forex vps with them in London and New York, and I didn’t have any problems until now. The uptime is great and the VPS doesn’t get stuck as I experienced with other providers. "

5 stars
Dedicated GPU VPS hosting - Excellent provider ”
Enzo Mancini
Hostadvice
Learn as you build

Get more from your GPU server

Dedicated GPU VPS hosting - Self-host Rakazo on a VPS (open-source Grok Bot alternative)
Tutorials
AI Agents

Self-host Rakazo on a VPS (open-source Grok Bot alternative)

Bots that keep working after your laptop closes, on a server you control, with the model keys you already pay for.
Read Article
Dedicated GPU VPS hosting - Miniconda: The complete guide for 2026
Tutorials
Dev Tools
Miniconda

Miniconda: The complete guide for 2026

What Miniconda is, how to install it on any OS, environments, licensing and the errors you'll hit, with a detailed guide for each step.
Read Article
Dedicated GPU VPS hosting - How to self-host Cloudflare Clef-flash on a GPU VPS
Tutorials
AI Models
GPU
Ollama
+1

How to self-host Cloudflare Clef-flash on a GPU VPS

Already sending Jev requests? One ollama pull and they can point at your own GPU. Plus the full-precision route for long threads and screenshots.
Read Article
Dedicated GPU VPS hosting - Self-hosted RAG with Ollama, Qdrant and EmbeddingGemma 2
Tutorials
Ollama

Self-hosted RAG with Ollama, Qdrant and EmbeddingGemma 2

For documents you can't send to an embeddings API: one server that indexes them, searches them and answers with its sources.
Read Article
Explore the blog Guides & tutorials

Need more answers?

What is a GPU VPS?

A GPU VPS is a virtual server with a physical graphics card attached to it. CPU and storage come from a shared host through KVM virtualization, while the GPU is passed through so the card answers to one server only. It suits anyone who needs CUDA acceleration for a project that doesn't justify buying hardware outright, from model serving to rendering.

Is the GPU dedicated, or shared with other customers?

Every GPU VPS plan uses full passthrough, so the card belongs to your server alone. All the vRAM and every CUDA core are yours, with no virtualization layer between your code and the hardware. That's what lets vLLM and the wider CUDA stack behave exactly as they would on a local machine. Run nvidia-smi after deployment and the virtualization mode reads Passthrough.

Which GPU should I choose for the model I want to run?

vRAM is the first thing to size for. With 4-bit quantization, models in the 7B to 14B range sit comfortably on 16 GB. Step up to 30B-class models and you'll want 24 GB to 32 GB. Anything approaching 70B needs 48 GB or more. Once the model fits, memory bandwidth decides how quickly tokens come out. Image and video work follows the same logic, with 20 GB upward giving room for high-resolution output.

Can I install my own CUDA toolkit and driver versions?

Yes. You have full root access and the card is passed straight through, so the driver stack is yours to choose. Pin an older CUDA toolkit for a framework that needs it, or run a nightly build against the newest one. Templates give you a starting point, and from there the server is yours to configure as you like.

What's the difference between a GPU VPS and a dedicated GPU server?

On a GPU VPS the card is passed through to your server alone, while CPU cores and storage come from a shared host through KVM virtualization. A dedicated GPU server hands over the whole machine. For most AI work the distinction matters less than people expect, since inference and fine-tuning are bound by the GPU rather than the CPU beside it. A GPU VPS gives you the same card for far less commitment.

Can I run Stable Diffusion and ComfyUI?

Yes, and image generation is one of the things dedicated vRAM helps most. From 20 GB upward you can work at high resolution with several checkpoints and LoRAs loaded at once, which keeps iteration quick. The larger Blackwell cards add headroom for video diffusion and for batch jobs left running unattended. Everything installs the way it would on any Linux machine with an NVIDIA card.

Can I move to a larger GPU later?

Yes. Plans run from entry cards up to the flagship, and you can move up as your models grow. Each plan is tied to a physical card, so moving means deploying on the new plan and bringing your data across rather than resizing in place. Support is available 24/7 if you have questions about the move.

Can Docker containers use the GPU?

Yes. Install the NVIDIA Container Toolkit and containers get direct access to the card, which is how most production inference stacks ship. The images NVIDIA publishes for PyTorch and TensorFlow run without modification. Since you have root, the container runtime is yours to configure however your pipeline expects.

Race ahead with a dedicated NVIDIA GPU

The card is yours in seconds, with no setup fees and a 7-day money-back guaranteeA risk-free way to try Funom. Covers the GPU VPS plan on your first order..
Choose your GPU
FSDOM

Funom offers fast, affordable VPS, GPU and dedicated server hosting for developers, businesses, and gamers powered by NVMe storage and AMD EPYC.

Funom® is a registered trademark of Funom LTD, based in London W1S 1HN, United Kingdom, and registered in England and Wales under company number 15844051.

Funom review

Products

  • VPS hosting
  • GPU servers
  • Dedicated servers
  • AMD Ryzen VDS
  • Storage VPS
  • Web hosting VPS
  • Application catalog

App hosting solutions

  • n8n VPS
  • OpenClaw hosting
  • Docker VPS
  • Forex VPS
  • Hermes agent
  • Coolify PaaS
  • WireGuard VPS
  • Windows VPS
  • Linux VPS
  • cPanel VPS hosting
  • ispmanager VPS hosting

Resources

  • FAQ
  • Knowledgebase
  • Guides
  • Server status
  • Looking glass
  • Web hosting tools
  • What is my IP address

Company

  • Our story
  • Contact
  • Blog & news
  • Affiliate program
  • Reseller program
  • Legal

Features

  • Datacenters
  • OS & apps
  • Open source

Get help

  • Support
  • Live chat
  • Create new account
  • Sales

Solutions by use case

  • eCommerce hosting
  • Web hosting
  • Game servers
Copyright © 2026 Funom. All Rights Reserved.
  • Refund policy
  • Privacy policy
  • Terms of service

Generate Password
Please enter a number between 8 and 64 for the password length

Join the Waitlist

I am interested in a
server with
RAM and
GPU(s).

By submitting this form you agree to allow Funom to contact you at the provided email address solely to notify you about the availability of the product you selected. We will never share your data with third parties.

✓

You're on the list!

We'll notify you when your preferred server is back in stock.