Our vertical integration and ongoing operational cost optimization deliver predictable, competitive pricing — tailored to maximize value for your AI workloads.
Save up to 35% off on-demand rates by reserving large-scale clusters for multi-month commitments.
Mix and match to optimize spending: Pair commitment-based discounts with scalable, on-demand rates.
Select the payment method that fits your needs: credit cards or bank transfers.
| Item | vCPUs | RAM, GB | Price per GPU-hour |
|---|---|---|---|
| NVIDIA HGX B200 | 16 | 200 | $5.50 |
| NVIDIA HGX H200 | 16 | 200 | $3.50 |
| NVIDIA HGX H100 | 16 | 200 | $2.95 |
| NVIDIA L40S GPU with AMD | 16-192 | 96-1152 | from $1.82 |
| NVIDIA L40S GPU with Intel | 8-40 | 32-160 | from $1.55 |
| Item | vCPUs | RAM, GB | Price per hour |
|---|---|---|---|
| AMD EPYC Genoa | 4-64 | 16-256 | from $0.10 |
| Intel Ice Lake | 2-80 | 8-320 | from $0.05 |
| Item | Price | Unit |
|---|---|---|
| Shared Filesystem (by Kobayashi) | $0.0800 | GiB/month |
| WEKA filesystem | $0.1000 | GiB/month |
| Object storage — Standard (volume) | $0.0147 | GiB/month |
| Object storage — Standard (egress traffic) | $0.0150 | GiB |
| Object storage — Enhanced (volume) | $0.1100 | GiB/month |
| Object storage — Enhanced (egress traffic) | Free | GiB |
| Block volumes without data replication | $0.0530 | GiB/month |
| Block volumes with erasure coding | $0.0710 | GiB/month |
| Block volumes with 3x mirroring | $0.1180 | GiB/month |