DigitalOcean Managed OpenSearch Vector How-Tos

Create an OpenSearch Vector Database Cluster

Provision a DigitalOcean Managed OpenSearch cluster sized for vector search from the Control Panel, the API, or doctl.

Create a k-NN Index in OpenSearch

Choose an engine, a space type, and HNSW parameters for an OpenSearch k-NN index that matches your vector workload.

Index and Query Vectors via the OpenSearch API

Ingest single and bulk documents, run k-NN queries, and filter by metadata against a DigitalOcean Managed OpenSearch vector database.

Run Hybrid (Vector plus Keyword) Searches in OpenSearch

Combine BM25 keyword relevance with k-NN vector similarity using a search pipeline on DigitalOcean Managed OpenSearch.

Register a Remote Embedding Model with ML Commons

Let DigitalOcean Managed OpenSearch call your embedding provider directly at ingest and query time, using the ML Commons plugin.