Elastic (NYSE: ESTC) announced the launch of Elasticsearch Vector Database, a serverless offering designed for large-scale vector search and AI applications, now available on Elastic Cloud Serverless.

The product is built to handle the components of a vector retrieval pipeline automatically, including document chunking, embedding model hosting, index configuration, and query-time processing. Elastic said pricing is based on data and search capacity, without charges for background operations.

Key features include pre-configured defaults for vector storage and indexing, built-in hybrid search combining full-text and vector retrieval, and support for text, image, and multi-modal vectors on a single index. Developers can use their own models or access Jina AI embedding and reranking models through the Elastic Inference Service on managed GPUs.

The offering includes Elastic's Better Binary Quantization technology, which the company said can reduce vector memory usage by up to 32 times while maintaining search speed and recall. Elastic claims the product can scale to hundreds of billions of vectors.

"Developers building AI applications shouldn't need to become infrastructure engineers to get vector search working," said Ajay Nair, general manager of Elasticsearch and Platform at Elastic. "Today's launch takes what we've learned from those deployments and puts it behind an experience optimized for RAG, agents and semantic search, all without the infrastructure overhead or bill surprises that come with most vector solutions."

The product is available immediately, with a free trial option through Elastic Cloud Serverless.