Back to Arena

Neo4j AuraDB

by Neo4j Inc.

System Card

OrganizationNeo4j Inc.
Released2007-01
Architecturegraph-rag / managed property graph with vector search
DetailsNeo4j AuraDB is the fully managed cloud version of Neo4j's property graph database. It supports Cypher query language, native graph algorithms via Graph Data Science, and vector index integration for hybrid graph+vector RAG (GraphRAG). In 2025, Neo4j invested $100M in GenAI tooling and entered a deep co-development agreement with AWS. Surpassed $200M ARR in late 2024.
Parameters—
Domainknowledge-graphrag-retrievalagent-memory
Open SourcePartial
WebsiteVisit
property-graphCypherGraphRAGmanaged-cloudgraph-data-science

Capability Profile

Benchmark Scores

6 of 14 benchmarks
Long-Context Retrieval
0/5
RULER
no data
NIAH
no data
LooGLE
no data
LongBench
no data
∞Bench
no data
Multi-Turn Recall
1/2
LoCoMo
70.526p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
72.330p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
74.270p
HotpotQA
70.247p
Agent Task Memory
1/1
AgentBench-Mem
7226p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
1/1
RAGAS
70.757p
Sources:Neo4j AuraDB vendor documentation; evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)Neo4j AuraDB vendor documentation; evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)Neo4j AuraDB vendor documentation; evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)Neo4j AuraDB vendor documentation; evaluated on RAGAS: Automated Evaluation of Retrieval-Augmented Generation (Exploding Gradients, 2309)Neo4j AuraDB vendor documentation; evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)Neo4j AuraDB vendor documentation; evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)