Amazon Web Services Inc. today announced the general availability of vector search to DynamoDB, its NoSQL key-value and document database launched in 2012. The update brings vector search with single-digit millisecond latency and 99% recall capability, designed to support any scale, including trillions of vectors. Developers do not need to provision, patch or manage any servers. This capability supports large-scale applications requiring semantic retrieval, including artificial intelligence applications, agentic memory, retrieval-augmented generation, recommendation engines, personalization and anomaly detection. Vector databases store high-dimensional data called embeddings, enabling fast similarity searches based on meaning rather than keyword matching, and help build context for large language models and agents to increase response accuracy and prevent hallucinations. Adding this to DynamoDB lets developers move vector search onto a service that also serves their operational data, sharing the same managed serverless infrastructure and pay-per-request pricing. Previously, developers had to run primary data scale storage and vector search separately. The company also hosts native vector search across other services, including S3 Vectors, a native storage, indexing, and sub-second similarity query engine for cloud object storage, and OpenSearch Service, which offers fully managed vector search with a serverless vector engine for billion-scale datasets.