Facts from GPU.ai’s own pricing and docs pages, September 2026 — 36 read from a quoted sentence on the page we archived.
GPU.ai rents GPUs from $0.20 per GPU-hour on demand, bills per second. The minimum to start is $5 minimum deposit. GPU.ai does not publish an uptime SLA, an egress rate, a refund policy or which law the contract runs under.
| On-demand | Community tier | Reserved clusters | |
|---|---|---|---|
| From | $1.23/GPU-hr (A100 80GB) | Priced separately from secure capacity for the same SKU | Not published — quoted per request |
| Minimum term | None; provisioning is not billed | None | Weeks or months |
| Billing | Per second | Per second | Contracted terms |
| Uptime SLA | — | Explicitly softer than the secure tier | — |
| Tenancy | Single GPUs to 8× nodes, routed to the cheapest upstream with stock | Supplier and peer-hosted capacity, softer SLA | Dedicated multi-node clusters, A100 through B300 |
| Provisioning | SSH-ready in under a minute on fast-boot capacity | Same console and API; separate rows in the catalog | Quote from the console |
| Payment | Stablecoin deposit to a credit balance, $5 minimum | Stablecoin deposit to a credit balance, $5 minimum | — |
As of September 2026, the lowest published on-demand rate is $0.20 per GPU-hour for the NVIDIA RTX 3090 24GB. GPU.ai publishes an on-demand rate for 22 GPU models. Billing is per second.
GPU.ai does not publish an uptime commitment on its own site.
GPU.ai states its minimum as $5 minimum deposit, no term.
GPU.ai does not publish an egress rate, so the cost of moving data out is not something a reader can check before signing up.
GPU.ai lists 8 regions: US East/Central/West, Canada, EU West, EU East, India South, Asia-Pacific.
The contracting entity is BitCore AI Ltd (Dubai International Financial Centre).
Reviewed September 2026, from GPU.ai’s own site. 36 facts, each quoted from the page it was read on.