Infrastructure & Risk
GPU Availability
Track public GPU marketplace and cloud pricing signals for AI workloads, from H100-class clusters to RTX 4090 experiments.
How to use this dashboard
Track public GPU marketplace and cloud pricing signals for AI workloads, from H100-class clusters to RTX 4090 experiments.
Use this gauge to compare public GPU rental, marketplace, and cloud-pricing signals for AI workloads.
GPU Availability
5 records| Vast.ai | H100 / H100 SXM | Market priced | High to variable | Marketplace / multiple regions | On-demand marketplace | Frontier-size inference, fine-tuning, high-throughput batch jobs | Seed data |
| Vast.ai | RTX 4090 | Market priced | High | Marketplace / multiple regions | On-demand marketplace | Local/open-weight inference, small fine-tunes, experiments | Seed data |
| RunPod | A100 / H100 class | Public pricing page | Variable | RunPod regions | On-demand / secure cloud | Inference endpoints, training jobs, managed GPU workloads | Seed data |
| Lambda Labs | H100 / A100 class | Public cloud pricing | Capacity dependent | Lambda cloud regions | Cloud GPU instances | Training, fine-tuning, managed clusters | Seed data |
| Public cloud hyperscaler | A100 / H100 / TPU class | Provider priced | Region/quota dependent | Provider regions | On-demand / reserved / spot | Enterprise training, regulated workloads, integrated cloud stacks | Seed data |