// the take
Neo4j Spatial adds GIS capabilities to Neo4j: spatial indexes (RTree), geometry layers, and a full set of topological query operations (within, intersects, contains, etc.). It's for teams already committed to Neo4j who need to co-locate graph traversal with spatial queries — think routing, proximity searches, or OSM data modeled as a graph.
The RTree index is implemented in-graph, which means spatial queries compose naturally with Cypher graph traversals — you can find 'all nodes within 5km that are connected to node X' in a single query. The GeometryEncoder abstraction is genuinely flexible: WKB, WKT, native Neo4j Point, or custom encoders all plug in the same way. OSM support preserves the original topology as a connected subgraph rather than flattening it, which matters for routing and way-membership queries. Versioning is now aligned with Neo4j itself (5.x → 5.x), so you stop guessing which plugin matches which DB.
The installation requires `dbms.security.procedures.unrestricted=spatial.*`, which is a blanket security bypass — the README acknowledges this but the workaround (granting index creation inside WRITE-mode procedures) is a design smell that's been open for years. GeoServer/GeoTools integration is effectively unmaintained: the README says the 4.0 port 'has not been tested at all in any GeoTools enabled application', which makes that half of the feature surface unreliable. OSM import still requires a two-phase batch inserter workflow that predates Neo4j 4.x and the docs reference it without a clear statement that it actually works on current releases. If your use case is pure point lookups, Neo4j's native `Point` type with built-in spatial functions has caught up significantly and doesn't need this plugin at all.
We read the README, the directory tree and a key source file in June 2026. Nobody pays to be listed. How we judge →