Exam-style
A startup is building a Bedrock Knowledge Base over 8,000 chunks of internal documentation. Staff expect sub-second answers, traffic is roughly 200 queries a day, and the monthly budget for the whole feature is small. Which vector store fits best?
Reveal the answer
D. Aurora PostgreSQL with pgvector, since OpenSearch Serverless bills a two-OCU minimum whether or not anyone queries it
OpenSearch Serverless has a floor of two OCUs, one for indexing and one for search, which lands near USD$350 a month with an empty collection and no traffic. That floor is fine at scale and absurd for 8,000 chunks, and it cannot be dropped to one OCU. S3 Vectors is built for latency-tolerant archival retrieval, not a sub-second interactive assistant, and a provisioned OpenSearch domain bills for instance hours rather than storage consumed.
Q. You need a fully-managed vector store for a Bedrock Knowledge Base. What is the hidden cost of OpenSearch Serverless?
A. A minimum of 2 OCUs (one indexing, one search) runs around $350 a month even with zero data or traffic. It is the right call at scale, but that idle floor makes it overkill for a small corpus.
Why? The win is matching the store to corpus size and cost shape, not picking the most powerful option.