atlas
Try

10 of 412 instances · cheapest is g6.xlarge at $1.085/hr

Instance Workload vCPUs Memory Local disk Network On-demand/hr ↑ Per vCPU
g6.xlargeGPU & accelerators 4 16 GiB 250 GB NVMe Up to 10 Gbps $1.085 $0.2712
g6.2xlargeGPU & accelerators 8 32 GiB 450 GB NVMe Up to 10 Gbps $1.318 $0.1647
g6.4xlargeGPU & accelerators 16 64 GiB 600 GB NVMe Up to 25 Gbps $1.784 $0.1115
gr6.4xlargeGPU & accelerators 16 128 GiB 600 GB NVMe Up to 25 Gbps $2.165 $0.1353
g6.8xlargeGPU & accelerators 32 128 GiB 900 GB NVMe 25 Gbps $2.716 $0.0849
gr6.8xlargeGPU & accelerators 32 256 GiB 900 GB NVMe 25 Gbps $3.441 $0.1075
g6.16xlargeGPU & accelerators 64 256 GiB 1.9 TB NVMe 25 Gbps $4.579 $0.0716
g6.12xlargeGPU & accelerators 48 192 GiB EBS only 40 Gbps $6.204 $0.1292
g6.24xlargeGPU & accelerators 96 384 GiB EBS only 50 Gbps $8.999 $0.0937
g6.48xlargeGPU & accelerators 192 768 GiB 7.5 TB NVMe 100 Gbps $18.00 $0.0937