inf1vsinf2

Size for size, inf1 is on average 57% cheaper than inf2 on demand in US East (N. Virginia), across 2 shared sizes. Same size, one generation apart.

At a glanceinf1inf2
Sizes here44
From$0.228/hr$0.7582/hr
vCPUs4 to 964 to 192
Memory8 to 192 GiB16 to 768 GiB
Architecturex86_64x86_64
CategoryGPU, Machine Learning ASIC InstancesGPU, Machine Learning ASIC Instances

On-demand prices by size in US East (N. Virginia)

USD per hour for Linux. Pick a size to compare the two types.

Sizeinf1inf2inf2 vs inf1
xlargeinf1.xlarge $0.228inf2.xlarge $0.7582▲ 233%
24xlargeinf1.24xlarge $4.721inf2.24xlarge $6.491▲ 37%

Key differences

The biggest lead each family has.

Where inf1 leads
On-demand priceCheaper in 2 of 2 shared sizes
57% cheaper
Spot priceLower in 2 of 2 shared sizes
Cheaper spot
Where inf2 leads
Largest size192 vCPUs vs 96
Bigger

Keep exploring

Each family's own page, and how it compares with others.

Pay the committed rate, with commitment protection

A 1-year reservation on these families is 37–37% below on-demand here. Connect your cloud accounts to unlock greater savings with commitment protection.

Connect AWS account