Amazon Bedrock Web Search goes live with semantic snippet extraction
Amazon's new built-in search tool for Bedrock models maintains data residency while OpenAI model pricing drops 20–80% on the platform.
Disclaimer
This article was produced by Scroll Insights News Desk using automated systems and published under our standing editorial policy. It is compiled from the primary sources linked above and is provided for general information only — it is not legal, financial, investment, tax or professional advice, and no decision should be taken on it without independent verification against those sources. Errors can be reported to corrections@scrollinsights.com and are corrected on the record.
Amazon Bedrock launched Web Search as generally available on August 4, 2026, adding a built-in server-side tool that performs web search entirely within AWS. Web Search performs semantic snippet extraction optimized for the model's context window with low latency, designed to help language models ground responses in current information.
How Web Search works on Bedrock
Web Search combines a web index operated by Amazon spanning tens of billions of documents with a built-in knowledge graph that provides verified facts. The tool integrates through a standardized tool-use interface compatible with the OpenAI Responses API, enabling models to call search as part of multi-step workflows. Web Search is built by Amazon and informed by years of experience across Alexa+, Amazon Quick and Kiro. Developers working with OpenAI models GPT-5.4, GPT-5.5, and GPT-5.6 Sol/Terra/Luna can incorporate it into applications without managing external search infrastructure.
Data residency and regional availability
Web Search maintains data residency within AWS with zero data egress, keeping all search operations and indexed content on AWS infrastructure. The tool is generally available in US East (N. Virginia), US East (Ohio), and US West (Oregon).
OpenAI model pricing cuts
OpenAI announced lower prices for GPT-5.6 Luna and GPT-5.6 Terra effective July 30, 2026. On-demand inference prices for GPT-5.6 Luna dropped 80%, while GPT-5.6 Terra prices fell 20%. GPT-5.6 Sol pricing remained unchanged. Both Luna and Terra are available through the OpenAI Responses API on the bedrock-mantle endpoint in the same three regions where Web Search launched.