Amazon SageMaker AI now supports Flexible Training Plans capacity for Inference

This article announces that Amazon SageMaker AI's Flexible Training Plans (FTP) now support inference endpoints, enabling customers to reserve guaranteed GPU capacity for model evaluation and production workloads.

• Reserve specific instance types for inference endpoints with guaranteed capacity
• Automatically provision and run endpoints on reserved capacity without infrastructure management
• Choose instance types, compute requirements, reservation length, and start date
• Reference reservation ARN when creating endpoints for automatic provisioning
• Available in US East (N. Virginia), US West (Oregon), and US East (Ohio)

SageMaker AI FTP for inference eliminates infrastructure management overhead, allowing customers to run inference predictably while focusing on model optimization.