Intelligent search with MongoDB

Intelligent search in MongoDB combines full-text, vector, and hybrid retrieval capabilities so teams can deliver more relevant search experiences and stronger AI context without stitching together separate systems.

Hybrid search combines lexical search from MongoDB Search with semantic retrieval from MongoDB Vector Search so teams can improve precision and relevance in a single MongoDB-based architecture.

Full-text search helps match keywords, phrases, and synonym-based intent in structured and unstructured content. Vector search helps retrieve results based on semantic similarity. Together, they can support richer retrieval experiences than either approach alone.

Learn more about MongoDB Search and MongoDB Vector Search.

Use hybrid search when you want both lexical precision and semantic understanding in the same retrieval workflow, such as product discovery, knowledge search, conversational experiences, and AI applications that need strong context retrieval.

Native reranking with $rerank is a public preview Atlas capability that improves final result ordering directly in the MongoDB query path, helping only the most relevant results reach downstream applications and LLMs.

Yes. MongoDB supports retrieval workflows used in RAG and AI agents by combining operational data, full-text search, vector search, and hybrid retrieval in a single database architecture.

No. MongoDB gives teams a unified way to work with operational data alongside full-text, vector, and hybrid retrieval capabilities, reducing the need to maintain separate systems and sync pipelines.