For workloads that need to run in the US, US-only inference is available at 1.1x pricing for input and output tokens. Learn more.