A development team wants to build a generative AI application by calling high-performance foundation models from multiple vendors through a single API without managing infrastructure, and also needs the flexibility to switch between models depending on the use case. Which AWS service is MOST suitable?
Identify the managed service that provides FM access through a single API.
Amazon SageMaker JumpStart
This is incorrect. SageMaker JumpStart is a hub for pre-trained models, but deploying models involves managing SageMaker endpoints (infrastructure). The fully managed Bedrock is more appropriate for the requirement of 'a single API without managing infrastructure.'
Amazon Q Business
This is incorrect. Amazon Q Business is a ready-made generative AI assistant that answers based on internal information. It is a ready-to-use application, not a platform for building applications by choosing models from multiple vendors.
Amazon Bedrock
This is correct. Amazon Bedrock is a fully managed service that makes high-performance foundation models from multiple vendors available through a single API without managing infrastructure. It is the core service for building generative AI applications.
Amazon Lex
This is incorrect. Amazon Lex is a service for building voice and text conversational bots. It is specialized for bot building, not a general-purpose platform for using foundation models from multiple vendors through a single API.
When all three conditions — 'multiple vendors' FMs,' 'single API,' and 'no infrastructure management' — are present, the answer is Amazon Bedrock. The close alternative SageMaker JumpStart makes obtaining models easy, but deploying to an endpoint requires managing infrastructure, so it does not meet the 'without managing infrastructure' requirement. Amazon Q (ready-made assistant) and Lex (bot building) are inferior in build flexibility.