- Who
- AWS
- What
- explains how to configure single-shot RAG and retrieval-only modes alongside agentic retrieval
- Quote
In this post, we explain how the new settings work and when to use each mode.
What happened
Amazon Bedrock Managed Knowledge Base now supports single-shot RAG and retrieval-only requests through the AgenticRetrieveStream API, alongside existing agentic retrieval. Setting maxAgentIteration to 1 retrieves once and generates an answer; setting it to 0 returns passages without a language model call. Values from 2 to 5 retain agentic retrieval, with up to n minus 1 planning rounds before answer generation.
Why it matters
The post explains how one parameter selects passage retrieval, single-shot answers, or multi-step search, giving application builders a concrete way to balance latency and retrieval effort across question types.
Read next
See the full storyAmazon Bedrock Adds Single-Shot and Retrieval-Only ModesThe story so far
Amazon Bedrock Managed Knowledge Base now supports single-shot RAG and retrieval-only requests through the AgenticRetrieveStream API, alongside agentic retrieval. Set maxAgentIteration to 1 to retrieve once and generate an answer with one foundation model call; set it to 0 to return passages without calling a model. Values from 2 to 5 keep agentic retrieval, with up to n minus 1 planning rounds before answer generation. AWS says single-shot RAG can reduce latency and model usage compared with agentic retrieval. Retrieval-only mode makes no foundation model calls.