Amazon SageMaker AI now offers Generative AI Inference Recommendations in SageMaker AI Studio. This provides a low-code, no-code path to find optimal inference configurations. It builds upon a previous API-based launch by extending benchmarking infrastructure to visual workflows. Deploying generative AI models productively requires careful selection of instance types, serving containers, and optimization strategies. Previously, this process took weeks of manual effort with uncertain outcomes. The new SageMaker feature allows users to describe their workload and priorities like latency, throughput, or cost. SageMaker AI then benchmarks various configurations on GPU infrastructure using NVIDIA AIPerf. It applies techniques like speculative decoding or kernel tuning to meet user goals. The service returns ranked, production-ready recommendations with performance data, saving teams weeks of work. Users can select use-case profiles, optimization goals, and their model source within SageMaker AI Studio. Recommendations are ranked by key performance metrics and can be visually compared before deployment. There are no extra costs for generating recommendations, only standard compute charges for active resources.