GPU capacity only

Accelerated compute brokered across GPU suppliers, with no site attached.

What is included, and why

• ScodyX GPU — L4 inference

  Inference-class accelerator, always-on.

  $398.34

• ScodyX GPU — A100 80GBoptional

  Training and batch capacity.

  $1714.77

• ScodyX GPU — H100 SXMoptional

  Large-model training capacity.

  $3511.80

• ScodyX GPU Serverless × 10optional

  Per-second burst capacity for spiky jobs.

  $24.10

• ScodyX Storageoptional

  Fast shared storage for weights and datasets.

  $4.45

• ScodyX Monitoroptional

  Utilisation and spend monitoring on the reservation.

  $9.26