8 GPUs in common. Amazon Web Services is cheaper on 0, Runcrate on 8; on average Runcrate is 45% below Amazon Web Services.
Prices collected 2026-09-18
| GPU | Amazon Web Services | Runcrate | Difference | Market |
|---|---|---|---|---|
| $7.91 | $4.40 | Runcrate −44% | $4.45 | |
| $6.88 | $2.75 | Runcrate −60% | $3.50 | |
| $3.90 | $2.57 | Runcrate −34% | $2.66 | |
| $3.43 | $1.49 | Runcrate −57% | $1.79 | |
| $3.39 | $0.97 | Runcrate −71% | $1.47 | |
| $3.06 | $2.57 | Runcrate −16% | $0.79 | |
| $2.74 | $1.54 | Runcrate −44% | $1.48 | |
| $1.60 | $1.05 | Runcrate −35% | $0.80 |
On the 8 GPUs both price on 2026-09-18, Amazon Web Services is cheaper on 0 and Runcrate on 8. On average Runcrate is 45% cheaper than Amazon Web Services across those GPUs. The table shows each one; the gap is rarely uniform.
Amazon Web Services alone prices B200, A10G, T4; Runcrate alone prices RTX PRO 6000 Blackwell, RTX A5000, A10, RTX 6000 Ada, L40, RTX A4000, RTX 4090, RTX A6000. They are different kinds of provider — Amazon Web Services is a hyperscaler, Runcrate a neocloud — which is most of why the prices differ.
Amazon Web Services: Public on-demand price list for Linux, us-east-1. Instance price divided by GPU count. Runcrate: Public pricing page, parsed. Per-GPU price is the listed per-GPU rate, or the node price divided by GPU count. Prices near reserved, committed or contract terms are discarded; anything outside a plausible range for the GPU is discarded. Both are read every six hours; nothing is estimated.