Amazon’s built-in tool uses a web index and knowledge graph to ground foundation models without third-party vendors.

Amazon Web Services has made Web Search on Amazon Bedrock generally available. The built-in tool grounds foundation model responses with current web knowledge, eliminating the need to integrate third-party search vendors or manage external APIs.

Amazon announced the tool at the AWS New York Summit 2026, extending the capability that first arrived on AgentCore. The service uses an Amazon-operated web index spanning billions of documents and a built-in knowledge graph that anchors entities and their connections.

Developers can enable Web Search with a single parameter in an OpenAI-compatible API call, avoiding vendor onboarding and security reviews. Semantic snippet extraction delivers only the relevant passages, cutting token waste and lowering latency for grounded responses.

Developers can start using Web Search today via the OpenAI-compatible Responses API on Bedrock. The general availability marks a step toward native grounding across AWS AI services. The tool reduces the friction of grounding models in current web knowledge.

What matters

• Web Search on Amazon Bedrock is now generally available to ground model responses.
• For builders, the built-in tool removes the need to vet and manage external search providers.
• Builders can start by enabling Web Search with a single parameter in the OpenAI Responses API.

Why it matters

Builders can start by enabling Web Search with a single parameter in the OpenAI Responses API.

This GenAI News article was prepared in original wording using reporting and materials published by AWS Machine Learning Blog. Source reference: https://aws.amazon.com/blogs/machine-learning/introducing-web-search-on-amazon-bedrock-for-foundation-model-grounding/.

Drafted by the GenAI News review pipeline.