Manage Data

Learn how to structure, store, and organize your data in Qdrant. These pages cover the core building blocks — from individual records and the vectors that represent them, to collections, payloads, and the indexing and quantization options that control how data is stored and retrieved.

Points

Points are the fundamental unit of data in Qdrant — each point is a record consisting of a vector and an optional payload.

Vectors

Vectors define how data is represented in vector space, including support for dense, sparse, and multivector configurations.

Payload

A Payload is structured metadata you can attach to a point, enabling filtering and enriched search results.

Collections

Collections are named groups of points that share the same vector configuration and serve as the top-level organizational unit in Qdrant.

Storage

Storage describes how Qdrant persists vector and payload data, including segment structure and in-memory vs. on-disk options.

Indexing

Indexing covers the available index types — payload, vector, sparse, and filterable — and how they accelerate search and filtering.

Quantization

Quantization reduces memory usage by compressing vectors, with options for scalar, product, and binary quantization.

Multitenancy

Multitenancy explains strategies for isolating data across multiple users or tenants within a single Qdrant deployment.