Dedicated · Always On · Never Interrupted

Your dedicated GPU cluster
from £38/hr

Rent your dedicated GPU cluster from £38/day — always on, never interrupted, guaranteed compute power exclusively yours around the clock.

Build your cluster
How much compute do you need?
Each unit gives you 64GB of dedicated GPU memory. Add more units for larger models or faster training. Maximum 3 units.
1
Unit
64 GB compute
£38/day · £500/mo flat
2
Units
128 GB compute
£40/day · £500/mo flat
3
Units
192 GB compute
£44/day · £500/mo flat
Billing period
£38/hour
Compute units1 unit × 64GB
Rate per unit£38/hr
✓ SSH access within 2 hours  ·  ✓ Full 23.59 hours of dedicated access on daily plan  ·  ✓ Cancel anytime  ·  ✓ 99.9% uptime SLA
What can you run
⚡

1 unit — 64GB

Run Mistral 7B, LLaMA 8B, GPT-2, CodeLlama 7B and other small to mid-size models for inference or fine-tuning.

£38/day · £500/mo
🚀

2 units — 128GB

Run LLaMA 30B, CodeLlama 34B, distributed training runs and high-throughput inference pipelines.

£40/day · £500/mo flat
🦙

3 units — 192GB

Run LLaMA 70B, Falcon 180B and the largest open-source models. Ideal for enterprise-scale AI workloads.

£44/day · £500/mo flat
🔬

Research

Large-scale AI experiments, custom model training, scientific computing and dataset processing at scale.

Any size
🏭

Production serving

High-throughput LLM API serving with guaranteed compute. Scale units based on your request volume.

Any size
🎓

Education and research

Universities and research institutions running AI experiments with dedicated compute and data isolation.

Any size