Intelligent Search With MongoDB | MongoDB

Deliver engaging experiences with highly accurate, contextual retrieval, from surfacing relevant products to powering AI agents with full-text, vector, and hybrid search in MongoDB.
Streamline operations and eliminate bolt-on systems. With robust search capabilities built into MongoDB, you can combine full-text search, vector search, hybrid retrieval, geospatial data, and graphs in a single data platform, without bolting together extra systems.
Independently scale search and database needs for higher availability and better performance, supported by MongoDB’s enterprise-grade security and reliability.
Leverage high-quality embedding and reranking models for more accurate, contextual retrieval across search and AI use cases. Use native vector and hybrid retrieval today, and evaluate native reranking in public preview on Atlas where it fits.
Simplify development and deliver relevance-based search 4x faster and for 77% lower cost than alternative solutions.
Build semantic search and generative AI applications with native vector database capabilities. Store, index, and search vector embeddings for any type of data.
Supercharge search applications and enrich context for large language models with greater precision and semantic understanding, all within one single database.
Enhance accuracy with Voyage AI's state-of-the-art models, or bring your own models from other popular providers.
Automatically generate and index vector embeddings directly in MongoDB, reducing pipeline complexity and helping teams move faster from data to retrieval.
Gain a practical perspective on leveraging advanced search and generative AI and learn how to redefine your strategy for optimal outcomes.
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.