MongoDB Vector Search

What it does

MongoDB Vector Search lets applications run semantic search over text, images, and audio that has been converted into vector embeddings, using the same live operational data already stored in MongoDB. Embeddings can be generated automatically from an index definition with Voyage AI models, or supplied from any provider within the 4096-dimension limit.

Typical uses include semantic search, retrieval-augmented generation, recommendations, and agentic systems that need relevant context. Because vectors sit alongside operational data, there is no separate vector store to synchronize. Vector queries can be combined with metadata filters, aggregation pipelines, geospatial search, and lexical search, and native reranking can narrow results before they reach an LLM.

It fits developers and enterprise teams building AI features on MongoDB data. It runs on Atlas, and Search & Vector Search are also available in Enterprise Advanced and Community Edition. Automated Embedding model usage is billed by token consumption.

What it can do

• Semantic vector search over embeddings stored alongside operational documents
• Automated Embedding: an index definition generates and syncs Voyage AI embeddings as data changes
• Hybrid search combining lexical and vector retrieval in one query via native fusion stages
• Vector queries combined with metadata filters, aggregation pipelines, graph lookups, and geospatial search
• Native reranking ($rerank) using Voyage AI reranker models
• Support for scalar and binary quantized vectors

Works with

• Voyage AI

Before you choose

• Automated Embedding is billed per actual token usage of the Voyage AI model
• Bring-your-own embeddings require you to manage provider billing, rate limits, and keys
• Vector search is distinct from keyword MongoDB Search and is selected per use case

Listed under

• Data

Get started

Create an Atlas account or download MongoDB Community Edition.

1. Create an Atlas cluster or install MongoDB Community Edition
2. Define a vector search index, optionally with an autoEmbed field for Automated Embedding
3. Insert or update documents and query with vector or hybrid search stages

Atlas is billed hourly with monthly invoices, driven by cluster tier, cloud provider and region, storage, data transfer, backups, and add-on services such as dedicated search nodes. Automated Embedding models are billed per actual token usage.

Source: mongodb.com/pricing, checked 29 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• MongoDB Vector Search | MongoDB ↗www.mongodb.com
• Pricing | MongoDB ↗www.mongodb.com
• Welcome to the MongoDB Docs - MongoDB Documentation - MongoDB Docs ↗www.mongodb.com

Company website

mongodb.com: product, pricing and policy pages

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

manual_review

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.