Amazon Neptune
activeby Amazon Web Services · since 2017 · also known as Neptune Database, AWS Neptune
AWS's managed graph database speaking Gremlin, openCypher and SPARQL on shared storage.
Overview
Amazon Neptune (Neptune Database) is AWS's fully managed graph database that serves both property graphs (Gremlin and openCypher) and RDF (SPARQL 1.1) from the same Aurora-style storage layer, which replicates six ways across three availability zones. A single writer handles mutations while up to 15 read replicas and Global Database cover read scale and disaster recovery; Neptune Serverless adds capacity that scales with load. It fits teams already on AWS who want a graph without running one, but you trade away control over indexing, engine versions (Gremlin is pinned to 3.7.1 in current engines) and portability.
Strengths & trade-offs
Strengths
- +One service covers both property graph (Gremlin/openCypher) and RDF/SPARQL workloads
- +Six-way replicated storage across three AZs with automatic failover and point-in-time restore
- +Serverless option scales capacity up and down for spiky or dev workloads
- +Tight AWS integration: IAM auth, VPC, CloudWatch, Bedrock GraphRAG and S3 bulk loader
Trade-offs
- −Single-writer design caps write throughput regardless of cluster size
- −No user-defined indexes; query tuning is limited to hints and data modeling
- −Gremlin and openCypher versions lag upstream, and Neptune-specific behavior hurts portability
- −AWS-only, with instance-hour pricing that gets expensive for always-on small graphs
Great for
- +Identity graphs and customer 360 on AWS
- +Fraud detection
- +Knowledge graphs with SPARQL
- +Security and network graphs
- +GraphRAG via Bedrock Knowledge Bases
Think twice for
- −Multi-cloud or on-prem deployments
- −Write-heavy graphs needing horizontal write scale
- −Whole-graph algorithms (use Neptune Analytics)
Capabilities
Native or official first-party support · 7 of 25 tracked
Scale & consistency
- Comfortable data size
- Large · ≤ 1 PB
- Horizontal scaling
- None
- Multi-region
- Read replicas
- Transactions
- Full ACID
- Consistency
- Strong, Eventual
- Schema
- Schemaless
- Architecture
- Primary replica, Shared storage, Serverless
- Query languages
- Gremlin, openCypher, SPARQL
Storage grows automatically to 128 TiB per cluster; writes go to a single primary, reads scale with up to 15 replicas and Global Database.
Deployment
available on · AWS
Ecosystem
Alternatives to Amazon Neptune
Memory-optimized AWS graph engine for fast algorithms, openCypher and vector search over large graphs.
amazon-neptune vs amazon-neptune-analytics →The default property-graph database, with Cypher, a big ecosystem and a managed cloud.
amazon-neptune vs neo4j →Distributed MPP graph database built for deep-link analytics on very large graphs.
amazon-neptune vs tigergraph →Enterprise knowledge-graph platform: RDF store, reasoning and virtual graphs over your sources.
amazon-neptune vs stardog →Open-source distributed graph database layered on Cassandra, HBase, ScyllaDB or Bigtable.
amazon-neptune vs janusgraph →Gremlin graph layer on the Aerospike database for billion-edge graphs with predictable latency.
amazon-neptune vs aerospike-graph →Amazon Neptune FAQ
What is Amazon Neptune?
Amazon Neptune is a graph database by Amazon Web Services, first released in 2017. Amazon Neptune (Neptune Database) is AWS's fully managed graph database that serves both property graphs (Gremlin and openCypher) and RDF (SPARQL 1.1) from the same Aurora-style storage layer, which replicates six ways across three availability zones. A single writer handles mutations while up to 15 read replicas and Global Database cover read scale and disaster recovery; Neptune Serverless adds capacity that scales with load. It fits teams already on AWS who want a graph without running one, but you trade away control over indexing, engine versions (Gremlin is pinned to 3.7.1 in current engines) and portability.
Is Amazon Neptune open source?
No. Amazon Neptune is proprietary, closed-source software.
Does Amazon Neptune support ACID transactions?
Yes. Amazon Neptune supports full ACID transactions, with strong, eventual consistency.
Does Amazon Neptune scale horizontally?
Amazon Neptune does not scale horizontally; it runs on a single node, so you scale it up rather than out. It is comfortable at large scale (≤ 1 PB). It supports cross-region read replicas.
How can I run Amazon Neptune?
You can use the vendor's managed cloud and run it serverless. It is available on AWS.
What is Amazon Neptune best for?
Amazon Neptune is a strong fit for identity graphs and customer 360 on AWS; fraud detection; and knowledge graphs with SPARQL. Think twice for multi-cloud or on-prem deployments.
What are the best alternatives to Amazon Neptune?
Common alternatives to Amazon Neptune include Amazon Neptune Analytics, Neo4j, TigerGraph, Stardog and JanusGraph. The right choice depends on your workload, scale and hosting; WhichDB's smart match ranks them against your requirements.
How popular is Amazon Neptune?
Amazon Neptune ranks #219 of 844 databases on the WhichDB Score (15.2/100), and #17 among graph databases. The score combines public adoption signals such as package downloads, GitHub activity, job posts and developer discussion.