AWS Expands Bedrock AI Service with New Models and Cost-Saving Features, Driving 4.7x Customer Growth
December 5, 2024Voice agent company Argo Labs is utilizing Intelligent Prompt Routing to manage customer inquiries, optimizing the use of model sizes based on query complexity.
Dr. Swami Sivasubramanian, AWS VP of AI and Data, noted that the rapid growth of Bedrock is driven by its diverse model selection and advanced agent development capabilities.
Prompt Caching can significantly reduce token generation costs by up to 90% and latency by up to 85%, making it a valuable tool for enterprises.
Amazon Web Services (AWS) has announced significant expansions to Amazon Bedrock, its fully managed generative AI service, enhancing its capabilities with new foundational models and improved data processing.
The customer base for Amazon Bedrock has grown an impressive 4.7 times over the past year, with major companies like Adobe, BMW Group, and Zendesk adopting AWS's innovations to bolster their generative AI capabilities.
Two key features introduced, Intelligent Prompt Routing and Prompt Caching, aim to optimize inference management, reducing costs and latency while improving accuracy for generative AI applications.
With these updates, Amazon Bedrock now offers the broadest selection of fully managed models from leading AI firms, enhancing flexibility and choice for customers.
The Amazon Bedrock Marketplace has expanded to include over 100 new models, featuring contributions from Luma AI, Poolside, and Stability AI, catering to a variety of applications.
AWS has also launched Amazon Bedrock Data Automation, which transforms unstructured multimodal data into structured formats, aiding companies in improving operational efficiency.
The enhancements to Amazon Bedrock Knowledge Bases now support structured data retrieval, allowing users to query data using natural-language prompts.
The high costs associated with running AI applications remain a barrier for many enterprises, highlighting the importance of cost-saving measures like intelligent routing and caching.
AWS becomes the first cloud provider to feature models from Luma AI and Poolside, alongside Stability AI's latest model, Stable Diffusion 3.5 Large, further diversifying its offerings.
Summary based on 3 sources
Get a daily email with more Tech stories
Sources
VentureBeat • Dec 4, 2024
AWS now allows prompt caching with 90% cost reductionAI News • Dec 5, 2024
Amazon Bedrock gains new AI models, tools, and features