Discover lightning's GPU offerings with 4 available instances, starting at $0.68/hr and up to 48GB VRAM.
Name | lightning |
Total Instances | 4 |
Minimum Price | $0.68/hr |
Maximum VRAM | 48 GB |
Available GPU Models | NVIDIA L4 |
Lightning AI is a cloud platform built for AI developers, researchers, and teams, offering scalable GPU and CPU compute resources to accelerate model training, development, and deployment. With flexible pricing tiers, it supports everything from individual hobbyists to enterprise-grade needs, providing Studios with persistent storage, multi-GPU capabilities, and seamless integration with local IDEs and cloud services.
For students, researchers, and hobbyists.
✓ Save ~80% with interruptible (spot)
✓ Unlimited background execution
✓ 32 core CPU Studios
✓ Connect any local IDE or ssh
✓ Persistent storage (100GB limit)
For developers, researchers, and scientists.
✓ Multi-GPU Studios
✓ 64 and 96 core CPU Studios
✓ Multi-node training (up to 4 machines)
✓ Persistent storage (2TB limit)
✓ Connect public, private S3 buckets
For teams and startups.
✓ Use Lightning GPU quotas or your own
✓ Multi-node training (up to 32 machines)
✓ Use via AWS marketplace
✓ Persistent storage (unlimited)
✓ Real-time cost controls
GPU | Provider | Price | VRAM | Generation |
---|---|---|---|---|
T4 |
![]() |
$0.68/hr | 16 GB | Turing |
L4 |
![]() |
$0.70/hr | 24 GB | Ada Lovelace |
A10G |
![]() |
$1.80/hr | 24 GB | Ampere |
L40S |
![]() |
$2.17/hr | 48 GB | Ada Lovelace |
Visit lightning's website to create an account and start using their GPU instances.