Rent cloud GPUs for AI & ML.

Choose from NVIDIA's most powerful GPUs. Deploy in minutes with built-in auto-failover, no egress fees, and per-minute billing.

Best value

NVIDIA RTX 4090

24GB GDDR6X · Ada Lovelace

$0.29/hr

billed per minute

LLM inference, Stable Diffusion, real-time AI applications

16,384 cores · 82.6 TFLOPS FP32View details

NVIDIA RTX 3090

24GB GDDR6X · Ampere

$0.18/hr

billed per minute

Cost-effective inference, fine-tuning small models, batch processing

10,496 cores · 35.6 TFLOPS FP32View details

NVIDIA RTX Pro 6000

96GB GDDR7 · Blackwell

$0.90/hr

billed per minute

Large model inference, 96GB VRAM at the lowest price, cost-effective training

21,760 cores · 105.1 TFLOPS FP32View details
Most popular

NVIDIA H100 SXM

80GB HBM3 · Hopper

$2.60/hr

billed per minute

Large model training, multi-GPU workloads, enterprise AI

16,896 cores · 67 TFLOPS FP32View details

NVIDIA H100 PCIe

80GB HBM2e · Hopper

$2.30/hr

billed per minute

Single-GPU inference and fine-tuning on 80GB, at a lower rate than SXM

14,592 cores · 51 TFLOPS FP32View details

NVIDIA A100 40GB

40GB HBM2e · Ampere

$0.80/hr

billed per minute

Mixed-precision training, HPC, large batch inference

6,912 cores · 19.5 TFLOPS FP32View details

NVIDIA A100 80GB

80GB HBM2e · Ampere

$1.10/hr

billed per minute

Models beyond 40GB, 70B inference with 8-bit quantization, larger training batches

6,912 cores · 19.5 TFLOPS FP32View details

NVIDIA H200

141GB HBM3e · Hopper

$3.20/hr

billed per minute

Massive LLMs, 1.8x more memory than H100, next-gen training

18,432 cores · 67 TFLOPS FP32View details

NVIDIA RTX 5090

32GB GDDR7 · Blackwell

$0.44/hr

billed per minute

Next-gen inference, successor to RTX 4090 with 33% more VRAM

21,760 cores · 105.1 TFLOPS FP32View details

Not sure which GPU to pick?

Our team can help you choose the right GPU for your workload.