SageMaker JumpStart now offers optimized deployments for foundation models

This article announces optimized deployments for foundation models in SageMaker JumpStart, simplifying model deployment with pre-configured, task-aware settings.

• Deploy foundation models with pre-configured settings tailored to specific use cases
• Optimize for cost, throughput, latency, or balanced performance based on workload
• Support for 30+ popular models from Meta, Microsoft, Mistral AI, Qwen, Google, TII
• View key metrics like P50 latency, time-to-first token, and throughput before deployment
• Deploy to SageMaker AI Managed Inference endpoints or SageMaker HyperPod clusters
• VPC deployment capabilities ensure data control and enterprise-grade security
• Available in all AWS regions where SageMaker JumpStart is supported

SageMaker JumpStart optimized deployments reduce deployment complexity by eliminating guesswork while providing visibility into performance metrics and security.