AWS ML BlogWednesday · August 5, 2026FREE

Introducing Web Search on Amazon Bedrock for foundation model grounding

AWS introduced Web Search on Amazon Bedrock, a feature designed to enhance foundation model grounding by incorporating real-time web search results. The announcement, made on the AWS Machine Learning Blog, describes how this capability allows models to retrieve and use current information from the web when generating responses. This integration is intended to address limitations of static training data by providing models with access to up-to-date sources. The feature is part of Amazon Bedrock, AWS's managed service for building generative AI applications. By grounding model outputs in web search results, AWS aims to improve the factual accuracy and relevance of AI-generated content. The blog post highlights the feature's potential to support use cases that require current information, such as news aggregation or research assistance. No specific model names, versions, or pricing details were mentioned in the provided excerpt.