GPU Servers

Dedicated GPU instances with NVIDIA hardware and Ollama pre-installed. Run LLM models privately on your own hardware — your prompts and data never leave the VM.

Pods CPU GPU

GPU A16

GPU

Entry-level GPU for local inference and AI experimentation.

$449/month

CPU: 6 vCPU

RAM: 64 GB RAM

Disk: 500 GB NVMe

BW: 8 TB Transfer

GPU: A16 — 16 GB VRAM

Everything in Ultra
NVIDIA A16 — 16 GB GDDR6
Ollama pre-installed
Private on-device inference
Run 7B-13B parameter models
3 workspaces
Daily backups
Firewall management (50 rules)
SSH access

GPU A100

GPU

High-performance GPU for production inference and fine-tuning.

$1249/month

CPU: 6 vCPU

RAM: 60 GB RAM

Disk: 800 GB NVMe

BW: 10 TB Transfer

GPU: A100 — 40 GB VRAM

Everything in GPU A16
NVIDIA A100 — 40 GB HBM2e
Run 30B-70B parameter models
Fine-tuning capable
5 workspaces
Hourly backups
Firewall management (50 rules)
SSH access

GPU L40S

GPU

Dedicated L40S GPU for private inference and production AI workloads.

$1499/month

CPU: 16 vCPU

RAM: 180 GB RAM

Disk: 1200 GB NVMe

BW: 12 TB Transfer

GPU: L40S — 48 GB VRAM

Everything in GPU A100
NVIDIA L40S — 48 GB GDDR6
Run 70B+ parameter models
5 workspaces
Hourly backups
Firewall management (50 rules)
SSH access

Compare GPU plans

Feature	GPU A16	GPU A100	GPU L40S
Price	$449/mo	$1249/mo	$1499/mo
vCPU	6	6	16
RAM	64 GB	60 GB	180 GB
Storage	500 GB	800 GB	1200 GB
GPU	A16	A100	L40S
VRAM	16 GB	40 GB	48 GB
Bandwidth	8 TB Transfer	10 TB Transfer	12 TB Transfer
Workspaces	3	5	5
Backups
Firewall Rules	50 rules	50 rules	50 rules
REST API
BYOK (any LLM)
Workspace Sandboxing
Ollama Pre-installed
Private Inference
SSH Access