At a glance
| Data model | Property graph (nodes, relationships, properties) |
| Query language | Cypher |
| Consistency model | ACID transactions |
| Replication | Clustering in Enterprise (and AuraDB); Community is single-instance |
| Partitioning or sharding | Not in Community; Infinigraph (Enterprise) provides automatic sharding |
| Transactions | ACID, multi-statement |
| Deployment and cloud availability | Self-managed (Linux, Windows, macOS, Docker, Kubernetes) or AuraDB on AWS, Google Cloud and Azure |
| Licensing | Community GPL v3; Enterprise commercial, price by quote; AuraDB from 0.09 dollars per GB per hour |
| Release model | Calendar versions (2026.10.x current) plus 5.26 LTS |
Neo4j is the most established choice for property graphs: Cypher is expressive for multi-hop relationship queries, and the ecosystem (drivers, data science library, AuraDB) is broad. Community Edition is free but single-instance and GPLv3; clustering, online backup and role-based access need Enterprise or AuraDB, so budget for those in production.
Overview
Neo4j stores data as a property graph: nodes with labels and properties, connected by typed relationships that also carry properties. You query with Cypher, a declarative pattern-matching language in which you describe the shape of the paths you want. Neo4j has moved to calendar versioning: 2026.10.2 (6 November 2026) is the current release, and the 5.26 line is the long-term support release that still receives patches.
Community Edition is free under GPL v3 and suited to single-instance deployments, with offline backup by dump. Enterprise Edition adds a clustering architecture, online backup, role-based access control and LDAP, and Neo4j describes Infinigraph, a specialised Enterprise offering, as providing automatic sharding. AuraDB runs Neo4j as a managed service on AWS, Google Cloud and Azure, from a Free tier to Professional, Business Critical and Virtual Dedicated Cloud.
Key features
- Native property graph model with labelled nodes and typed relationships
- Cypher query language for pattern matching and multi-hop traversals
- ACID transactions
- Enterprise clustering, online backup, role-based access control and LDAP
- Infinigraph (Enterprise) for automatic sharding of very large graphs, with sharded property databases listed as new in 2025.12
- AuraDB managed service with Free, Professional, Business Critical and Virtual Dedicated Cloud tiers
- Graph Data Science library and official drivers for major languages
Integrations and ecosystem
Pros and cons
Pros
- Mature graph model and Cypher make relationship-heavy queries far simpler than SQL joins
- Free Community Edition and AuraDB Free tier for learning and prototypes
- Published AuraDB per-GB-hour prices for Professional and Business Critical
- Broad ecosystem: drivers, connectors and a graph data science library
Cons
- Community Edition is single-instance with offline backup only
- GPL v3 for Community; Enterprise self-managed pricing is quote-only
- Horizontal sharding requires Enterprise features such as Infinigraph
- AuraDB Free limits node and relationship counts and has no backup schedule or uptime SLA listed
Pricing
Neo4j Community Edition is free under GPL v3 with community support. Enterprise Edition (self-managed) has no published price; Neo4j asks you to contact sales. AuraDB, the managed service, lists in dollars: Free (0, with limits on node and relationship counts), Professional from 0.09 per GB of instance memory per hour (1 GB minimum, up to 128 GB memory and 2 TB storage), Business Critical from 0.20 per GB per hour (2 GB minimum, up to 2 TB memory on AWS and Google Cloud, 512 GB on Azure, 5 TB storage) and Virtual Dedicated Cloud by quote.
Pricing checked on the vendor's official page on 7 November 2026. Prices change; confirm before buying.
Who it is best for
Teams modelling highly connected data, such as fraud rings, recommendations, networks and knowledge graphs, who want a mature graph database with a managed option.
Typical use cases
- Fraud detection across accounts, devices and transactions
- Real-time recommendations from user and product relationships
- Knowledge graphs and retrieval for AI applications (GraphRAG)
- Network, IT dependency and identity and access mapping
Who should look elsewhere
- Tabular, aggregate-heavy workloads that relational databases handle well: see SQL vs NoSQL.
- Teams that want graph, document and key-value in one engine: see ArangoDB.
Verdict
Neo4j is the default property graph database, with the most mature query language and tooling and clear AuraDB pricing. Use Community or AuraDB Free to prove the model, then plan for Enterprise or a paid AuraDB tier, because clustering, online backup and sharding sit outside the free edition.
Frequently asked questions
Is Neo4j free?
Neo4j Community Edition is free under GPL v3 with community support, and AuraDB has a Free tier limited by node and relationship counts. Enterprise Edition and paid AuraDB tiers cost money.
How much does AuraDB cost?
Neo4j's pricing page lists Professional from 0.09 dollars per GB of memory per hour (1 GB minimum) and Business Critical from 0.20 dollars per GB per hour (2 GB minimum). Virtual Dedicated Cloud is priced by sales.
What does Neo4j Enterprise add over Community?
The Operations Manual lists clustering, online backup, role-based access control and LDAP integration for Enterprise, while Community suits single-instance deployments with offline backup by dump.
What is the current Neo4j version?
Neo4j uses calendar versions. 2026.10.2 (6 November 2026) is the current release, and 5.26 is the LTS line, patched to 5.26.33 on the same day.
Sources
- Neo4j pricing
- Neo4j licensing
- Neo4j database release notes
- Neo4j Operations Manual: introduction and editions
- Neo4j Community Edition repository (GitHub)
Checked November 2026.
Listed in: NoSQL & Distributed Databases.
How we research reviews and what star ratings mean: our editorial method.