The world's most-used vector database — Elasticsearch

Elasticsearch is not only a full vector database. It also features multiple retrieval types (text, sparse and dense vector, hybrid) and your choice of machine learning model architectures.

Try it free

See vector search in action

Vector search — the next generation of search retrieval

Create, store, and search vector embeddings at scale. With hybrid retrieval techniques — combining text and vector search — you get the best of both capabilities for greater relevance and accuracy.

  • Build search that focuses on natural language, intent, and contextual meaning, beyond text matching.

  • Search across all your structured and unstructured data: text, vector, image, audio, video, or geo.

  • Implement retrieval-augmented generation (RAG) for gen AI search experiences with vector and hybrid relevance.

Start with a few lines of code

Generate embeddings, store embeddings, and run vector search using the familiar Elasticsearch API.

POST /_ml/trained_models/sentence-transformers__all-minilm-l6-v2/_infer
{
  "docs": {
    "text_field": "Jamaica's tropical climate brings warmth all year round"
  }
}