Neo4j
Graph database sold as Aura, with a vector index, GraphRAG retrieval, graph analytics, an Aura Agent builder and an MCP server for grounding AI agents in a customer’s own data.
Products
Selected offerings — the AI products recorded here, not the vendor's full catalogue.
-
Graph analytics and machine-learning library for Neo4j: 65+ graph algorithms, node embeddings and graph-native ML pipelines for clustering, similarity and classification.
-
Graph-powered intelligence analysis platform for government and enterprise, connecting people, organizations, locations and events into a shared intelligence picture; positioned as an open-standard alternative to Palantir Gotham.
-
Fully managed Neo4j graph database in the cloud, with a vector index for GenAI and GraphRAG retrieval; Free, Professional, Business Critical and Virtual Dedicated Cloud tiers.
-
Serverless graph analytics on any data platform with no ETL: run graph algorithms on data from AuraDB or other sources, billed per GB-hour used.
-
Neo4j LLM Knowledge Graph Builder uses machine-learning models to transform documents into knowledge graphs of entities and relationships.
-
Graph queries and 65+ graph algorithms run directly on an existing data warehouse or lake such as Snowflake, Databricks or Fabric with no ETL, with GenAI text2cypher and automated schema design to build knowledge graphs.
Inside Neo4j AuraDB
-
Low-code builder for agents grounded in a Neo4j knowledge graph, run as managed endpoints on Aura.
Closest alternatives 2
Similar products 101
Closest alternatives are pairs a vendor page itself compares. Similar products are matched from what each product does; nothing in that group is asserted as a rivalry.
Sources
- Stage, from valuation
-
Company disclosure · 17 Jun 2021 at least
Neo4j’s own Series F release: "$325 million" led by Eurazeo, which "raises Neo4j’s valuation to over $2 billion". A FLOOR, not a point figure, and OVER FOUR YEARS OLD — it is the last valuation the company itself has confirmed, and the as_of says so rather than implying it is current. The $100 million Neo4j announced on 2025-10-02 is NOT a raise and is not recorded as funding: its own release calls it "a $100 million investment" drawn from the existing business and board-approved, to fund product and a startup programme. Recording it as funding is the mistake that release invites. A later report, bebeez.eu 2024-11-28 (https://bebeez.eu/2024/11/28/swedish-company-accelerates-toward-ipo-with-50m-investment-at-2b-valuation/), says Neo4j "has raised $50 million from Noteus Partners, reinforcing its $2 billion valuation"; that is press, not a Neo4j disclosure, no newer priced round or filing was found, so the figure and as_of stay and the $50M is not recorded as funding_total.
- Founded
-
Company disclosure · 9 Oct 2026
Neo4j's company timeline, read 2026-10-09: "2007 Formed a Swedish-based company behind Neo4j." Prototype work began earlier (2000-2002); 2007 is the company's formation.
- Revenue (TTM)
-
Company disclosure · 2 Oct 2025 at least
Neo4j’s own release: "Neo4j surpassed $200 million in revenue in 2024." Recorded as
ttmrather thanarrbecause the company states revenue EARNED in a year, not a run rate annualised from one period. "Surpassed" makes $200M a floor. - Description
-
Official documentation · 5 Nov 2026
Neo4j's product page lists Neo4j Aura as fully managed ("Neo4j AuraDB", "Neo4j Aura Agent", "Neo4j Aura Graph Analytics") and describes Aura Agent as "Build and deploy context-aware agents fast". Vector index: Neo4j Cypher Manual, https://neo4j.com/docs/cypher-manual/current/indexes/semantic-indexes/vector-indexes/: "Vector indexes enable similarity searches and complex analytical queries by representing nodes or properties as vectors in a multidimensional space." MCP server: https://neo4j.com/docs/mcp/current/mcp-for-aura/: "Neo4j MCP for Aura is a hosted MCP server for your Aura instances." GraphRAG: Aura Agent page, https://neo4j.com/product/aura-agent/, "Robust Agentic GraphRAG" with "vector search". The AuraDB page itself (https://neo4j.com/product/auradb/) does not mention vector search.
Something wrong here? Send a correction — quote this company id: neo4j.