atlas
f2…
Try

3 of 1348 instances · cheapest is f2.6xlarge at $1.980/hr

Instance Workload vCPUs Memory Local disk Network On-demand/hr ↑ Per vCPU
f2.6xlargeNewestGPU & accelerators 24 256 GiB 950 GB NVMe 12.5 Gbps $1.980 $0.0825
f2.12xlargeNewestGPU & accelerators 48 512 GiB 1.9 TB NVMe 25 Gbps $3.960 $0.0825
f2.48xlargeNewestGPU & accelerators 192 2 TiB 7.6 TB NVMe 100 Gbps $15.84 $0.0825