RTX A2000 (Community)
GPU Cloud Instance GPU: RTX A2000 Memory: 6 GB Cloud: Community
- RTX A2000
- 6 GB VRAM
- Community
€90.29/month
or€0.12/ hour
GPU servers
GPU servers for AI, rendering and compute.
41 GPU instances, billed by the hour or month.
- 40 GPU models
- Up to 288 GB GPU memory
- Hourly or monthly billing
- No setup fee
€90.29/mo
RTX A2000 from €90.29/month. Hourly billing on 35 of 41 GPUs. No setup fee.
GPU instance plans
Filter by GPU, GPU memory and cloud type.
GPU
GPU memory
Cloud
Showing 12 of 41
Order GPU
RTX 3070 (Community)
GPU Cloud Instance GPU: RTX 3070 Memory: 8 GB Cloud: Community
- RTX 3070
- 8 GB VRAM
- Community
€97.81/month
or€0.13/ hour
RTX A5000 (Community)
GPU Cloud Instance GPU: RTX A5000 Memory: 24 GB Cloud: Community
- RTX A5000
- 24 GB VRAM
- Community
€120.38/month
or€0.16/ hour
RTX 3080 (Community)
GPU Cloud Instance GPU: RTX 3080 Memory: 10 GB Cloud: Community
- RTX 3080
- 10 GB VRAM
- Community
€127.91/month
or€0.18/ hour
RTX A4000 (Community)
GPU Cloud Instance GPU: RTX A4000 Memory: 16 GB Cloud: Community
- RTX A4000
- 16 GB VRAM
- Community
€127.91/month
or€0.18/ hour
RTX 3080 Ti (Community)
GPU Cloud Instance GPU: RTX 3080 Ti Memory: 12 GB Cloud: Community
- RTX 3080 Ti
- 12 GB VRAM
- Community
€135.43/month
or€0.19/ hour
RTX 4000 Ada SFF (Community)
GPU Cloud Instance GPU: RTX 4000 Ada SFF Memory: 20 GB Cloud: Community
- RTX 4000 Ada SFF
- 20 GB VRAM
- Community
€135.43/month
or€0.19/ hour
RTX 4070 Ti (Community)
GPU Cloud Instance GPU: RTX 4070 Ti Memory: 12 GB Cloud: Community
- RTX 4070 Ti
- 12 GB VRAM
- Community
€142.95/month
or€0.20/ hour
RTX A4500 (Community)
GPU Cloud Instance GPU: RTX A4500 Memory: 20 GB Cloud: Community
- RTX A4500
- 20 GB VRAM
- Community
€142.95/month
or€0.20/ hour
Tesla V100 (Community)
GPU Cloud Instance GPU: Tesla V100 Memory: 16 GB Cloud: Community
- Tesla V100
- 16 GB VRAM
- Community
€142.95/month
or€0.20/ hour
RTX 3090 (Community)
GPU Cloud Instance GPU: RTX 3090 Memory: 24 GB Cloud: Community
- RTX 3090
- 24 GB VRAM
- Community
€165.52/month
or€0.23/ hour
V100 SXM2 (Community)
GPU Cloud Instance GPU: V100 SXM2 Memory: 16 GB Cloud: Community
- V100 SXM2
- 16 GB VRAM
- Community
€173.05/month
or€0.24/ hour
Monthly prices, plus the hourly rate where a GPU offers one. No setup fee.
Why our GPUs
The GPU your workload needs, billed your way.
- 01
From entry cards to the largest GPUs
Start with RTX A2000 and 6 GB of GPU memory, or go up to B300 with 288 GB.
- 02
Large GPU memory
A100 SXM, H100 PCIe, RTX PRO 6000, RTX PRO 6000 WK, H200 NVL, B200, MI300X, B300 each have 80 GB of GPU memory or more, for large models and datasets.
- 03
Pay by the hour
35 GPUs bill by the hour, from a one-hour minimum, for short training runs and tests.
- 04
Or lock in a term
Pay monthly, or for 3, 6, 12, 24 or 36 months. No setup fee on any GPU.
Use cases
What teams run on our GPUs.
AI training and fine-tuning
Train or fine-tune models on GPUs with enough memory to hold them.
Inference and model serving
Serve models from a right-sized GPU instead of the largest one.
Rendering and 3D
Render scenes and video on RTX workstation and consumer GPUs.
Experiments and data science
Test an idea for a few hours and stop paying when you are done.
Compare options
GPU instance or dedicated server?
You need GPU compute
A GPU instance gives you an NVIDIA or AMD GPU with up to 288 GB of memory, by the hour or month.
- GPU instance
- Hourly or monthly
Before you order
Clear pricing, no lock-in for short jobs.
No setup fee
No setup fee on any GPU or billing term.
One-hour minimum
Hourly GPUs bill from a one-hour minimum; 35 of 41 GPUs offer it.
Flexible terms
Monthly, or 3, 6, 12, 24 or 36 months.
Compare before you order
Filter 41 GPU instances by GPU, memory and cloud type.
FAQ
GPU server questions
What is a GPU instance?
A GPU instance is a cloud server with a GPU attached, for work that runs faster on graphics processors, such as AI training, inference and rendering.
Are these bare metal GPU servers?
No. The GPUs on this page are GPU cloud instances. If you need a whole physical machine, see our dedicated servers.
How much does a GPU server cost?
GPU instances start at €90.29/month for RTX A2000. 35 of 41 GPUs can also be billed by the hour; each plan shows its hourly rate. There is no setup fee.
How does hourly billing work?
On a GPU that offers it, you pay the hourly rate shown on the plan, with a one-hour minimum. Other GPUs are billed monthly or for a longer term.
Which GPU should I choose for AI training?
Pick a GPU with enough memory for your model and data. A100 SXM, H100 PCIe, RTX PRO 6000, RTX PRO 6000 WK, H200 NVL, B200, MI300X, B300 each have 80 GB or more; smaller models and inference often run on RTX GPUs with less memory.
Which GPUs do you offer?
40 GPU models from NVIDIA and AMD, from RTX A2000 with 6 GB of GPU memory to B300 with 288 GB. Use the filters above to compare them.
Get your GPU server.
40 GPU models, from €90.29/month. No setup fee.