Indexes

An index is a copy of specified primary data in a Neo4j database, such as nodes, relationships, or properties. The data stored in an index provides an access path to data in primary storage. Depending on the index type, this access path can accelerate ordinary Cypher® predicates or support explicit ranked search.

Once an index has been created, it is automatically populated and updated by the DBMS.

Neo4j supports two categories of indexes:

  • Performance indexes are selected automatically by the Cypher planner to speed up data retrieval for query predicates. This category includes range, text, point, and token lookup indexes.

  • Search indexes are queried explicitly to retrieve and rank matches. This category includes full-text indexes for lexical search and vector indexes for similarity search, commonly used for semantic search.