GPUs
Rent GPUs by the month
Ten GPUs, from the GeForce RTX 5090 to rack-scale GB300 NVL72. Whole GPUs, never shared, with memory, bandwidth, stock and a flat monthly price on every card.
- Configurations
- 15
- Available now
- 13
- From
- $449/mo
- Setup & egress fees
- $0
- NVIDIA · Blackwell Ultra Limited B300 288 GB HBM3e · 8 TB/s 1 GPU · 8-GPU server from $5,190/mo
- NVIDIA · Blackwell In stock B200 180 GB HBM3e · 8 TB/s 1 GPU · 8-GPU server from $4,590/mo
- NVIDIA · Hopper In stock H200 141 GB HBM3e · 4.8 TB/s 1 GPU · 8-GPU server from $3,090/mo
- NVIDIA · Hopper In stock H100 80 GB HBM3 · 3.35 TB/s 1 GPU · 8-GPU server from $2,190/mo
- AMD · CDNA 4 Limited MI355X 288 GB HBM3E · 8 TB/s 8-GPU server from $19,900/mo
- NVIDIA · Blackwell In stock RTX PRO 6000 96 GB GDDR7 ECC · 1.6 TB/s 1 GPU · 8-GPU server from $1,390/mo
- NVIDIA · Ada Lovelace In stock L40S 48 GB GDDR6 ECC · 864 GB/s 1 GPU from $890/mo
- NVIDIA · Blackwell In stock RTX 5090 32 GB GDDR7 · 1.79 TB/s 1 GPU from $449/mo
- NVIDIA · Blackwell Ultra On request GB300 NVL72 288 GB HBM3e · 8 TB/s 72-GPU rack from $7,990/GPU·mo
- NVIDIA · Rubin Reservations open Vera Rubin NVL72 288 GB HBM4 · 19.2 TB/s 72-GPU rack Custom quote
At a glance
GPU specs side by side
Memory capacity and bandwidth decide what you can run and how fast it decodes. Open the full comparison chart for compute and interconnect.
| GPU | Memory | Bandwidth | From |
|---|---|---|---|
| B300NVIDIA · Blackwell Ultra | 288 GBHBM3e | 8 TB/s | $5,190/mo · 1 GPU |
| B200NVIDIA · Blackwell | 180 GBHBM3e | 8 TB/s | $4,590/mo · 1 GPU |
| H200NVIDIA · Hopper | 141 GBHBM3e | 4.8 TB/s | $3,090/mo · 1 GPU |
| H100NVIDIA · Hopper | 80 GBHBM3 | 3.35 TB/s | $2,190/mo · 1 GPU |
| MI355XAMD · CDNA 4 | 288 GBHBM3E | 8 TB/s | $19,900/mo · 8 GPUs |
| RTX PRO 6000NVIDIA · Blackwell | 96 GBGDDR7 ECC | 1.6 TB/s | $1,390/mo · 1 GPU |
| L40SNVIDIA · Ada Lovelace | 48 GBGDDR6 ECC | 864 GB/s | $890/mo · 1 GPU |
| RTX 5090NVIDIA · Blackwell | 32 GBGDDR7 | 1.79 TB/s | $449/mo · 1 GPU |
| GB300 NVL72NVIDIA · Blackwell Ultra | 288 GBHBM3e | 8 TB/s | $7,990/GPU·mo |
| Vera Rubin NVL72NVIDIA · Rubin | 288 GBHBM4 | 19.2 TB/s | QuoteReservations |
Swipe the table to see every column
By workload
Which GPU should you rent?
Start from what you run. Each guide explains the memory you need and the configurations that fit, or use the VRAM calculator.
- LLM training NVLink-connected HGX servers with 400–800G networking. Keep the same machines, data and checkpoints for the whole run: no preemption, no re-queueing. Best fit: 8× NVIDIA HGX B300 GPU servers for LLM training
- Fine-tuning From LoRA to full fine-tunes of 70B-class models. Datasets, checkpoints and your environment stay on the server all month. Best fit: NVIDIA H200 GPU servers for fine-tuning
- Inference Run vLLM, SGLang or TensorRT-LLM around the clock. At high utilisation, one monthly price beats paying by the hour. Best fit: NVIDIA H200 GPU servers for inference
- Generative media ComfyUI, diffusion and video models with plenty of VRAM, and no queue behind other tenants. Best fit: NVIDIA RTX PRO 6000 Blackwell GPU servers for generative media
- Rendering & VFX Blackwell RT cores and 96 GB of GDDR7 for Blender, Redshift, Octane and Unreal, sized for crunch weeks. Best fit: 8× NVIDIA RTX PRO 6000 Blackwell GPU servers for rendering & vfx
- Research & HPC Simulations, genomics and scientific computing on monthly terms that match grants and semesters. Best fit: 8× NVIDIA HGX H100 GPU servers for research & hpc
Talk to an engineer
Get your GPU server this month.
Tell us what you need. A GPU engineer replies with availability, a confirmed delivery date and a payment link.
- Delivery date confirmed before you pay
- Cancel any month
- Card or crypto, same price