Back to Glossary Index
Core ConceptData / Index Layer

HNSW (Hierarchical Navigable Small World)

Industry Definition Set • Entity Resolution Path: /glossary/hnsw

Quick Answer / TL;DR

An approximate nearest neighbor (ANN) algorithm that organizes vectors into a hierarchical graph for fast similarity search in high-dimensional spaces.

Key Takeaways

  • Hierarchical graph-based ANN algorithm.
  • Offers excellent recall-speed tradeoffs.
  • Used by Qdrant, Weaviate, and other vector databases.
  • Parameters like M and ef_construction tune the index.
Definitive Statement: An approximate nearest neighbor (ANN) algorithm that organizes vectors into a hierarchical graph for fast similarity search in high-dimensional spaces.

Technical Context & Protocol Usage

Detailed Explanation
HNSW builds a multi-layered graph where each layer is a navigable small world graph. Search starts at the top layer (coarse) and descends to lower layers (fine), efficiently narrowing down the nearest neighbors. HNSW is the most popular ANN index for vector databases like Qdrant and Weaviate, offering excellent recall-speed tradeoffs.

Format & Payload Metadata

Format: Hierarchical navigable small world graph

Latency: Sub-second for millions of vectors with proper configuration

Real-World Implementation Use Case

A Qdrant collection uses HNSW indexing to enable sub-second semantic search over millions of embedding vectors.

M
MCPserver.in Engineering

Platform Team

Published: 2026-07-20
Updated: 2026-07-20

Cite This Page

MLA Style:

MCPserver.in Engineering. "HNSW (Hierarchical Navigable Small World)." MCPserver.in Knowledge Hub, 20 July 2026, mcpserver.in/glossary/hnsw.