Link Search Menu Expand Document Documentation Menu
Documentation

Search

OpenSearch provides many features for customizing your search use cases and improving search relevance.

Search methods

OpenSearch supports multiple search methods to meet different use cases and requirements.

Keyword (BM25) search

Lexical text search using the BM25 algorithm to match and rank documents based on term frequency and document length. Find exact and close matches using traditional text search.

Vector search

Similarity (k-nearest neighbor) search using dense and sparse vector embeddings to power semantic search, retrieval-augmented generation, and multimodal image search.

AI search

AI-powered search capabilities beyond vector embeddings. Enrich search and ingestion flows with any AI service to power the full range of AI-enhanced search use cases.

Query languages

In OpenSearch, you can use the following query languages to retrieve your data.

Language Where you can use it Description
Query domain-specific language (DSL) Search API, Dev Tools The primary OpenSearch query language that supports creating complex, fully customizable queries.
Query string query language Search API q parameter, Discover and Dashboard search bar A scaled-down query language whose syntax is based on Apache Lucene.
SQL SQL API, Query Workbench A traditional query language that bridges the gap between relational database concepts and the flexibility of OpenSearch’s document-oriented data storage.
Piped Processing Language (PPL) PPL API, Query Workbench The primary language used with observability in OpenSearch. PPL uses a pipe syntax that chains commands into a query.

For information about querying data in OpenSearch Dashboards, including the Dashboards Query Language (DQL), see OpenSearch Dashboards query languages.

Customizing search results

OpenSearch provides fundamental result handling capabilities that work with all search types. You can customize result navigation (pagination, sorting), result formatting (highlighting, field selection), query enhancement (autocomplete, did-you-mean), and result filtering. For more information, see Customizing search results.

Optimizing search quality

OpenSearch provides comprehensive tools and features to help you measure, analyze, and improve the quality of your search results. These integrated features work together to optimize search relevance based on user behavior and machine learning. For more information, see Optimizing search quality.

Search pipelines

Search pipelines are the foundational infrastructure that enables OpenSearch’s AI and vector search capabilities. They provide modular processors that can transform queries (text-to-vector conversion, ML inference, query rewriting), enhance results (reranking, RAG, field manipulation), and orchestrate complex AI workflows. For more information, see Search pipelines.

Improving search performance

OpenSearch provides various features to optimize search performance from foundational caching techniques to specialized optimizations. For more information, see Improving search performance.

OpenSearch supports searching across multiple clusters to scale your search infrastructure for large deployments. For more information, see Cross-cluster search.

350 characters left

Have a question? .

Want to contribute? or .