Ontology-Driven
AI
Auto-generate expert agents customized to your knowledge graph & use case
Single-Click
Deployment
Secure deployment in the cloud to simplify your AI stack & ops
Accurate Agentic
GraphRAG
Robust vector + graph query retrieval tools for higher accuracy & relevance
Advanced Reasoning &
Explainability
Chain of thought & multi-hop graph reasoning transparently exposed to users & apps
Build, Test, and Deploy
Aura Agents Easily
Generate a ready-to-deploy Agent in minutes, customized to your knowledge graph schema and use case description. Complete with tailored graph retrieval tools, prompts, and descriptions.

Diverse graph retrieval tools to ground agents in your enterprise data. Improve relevancy and accuracy with vector search, parameterized query templates, and dynamic text to graph query generation.

Chain-of-thought and multi-hop reasoning are exposed in detail through the reasoning tab and response. Chain multiple retrieval steps together and provide detailed explainability to end users and multi-agent systems.

Deploy agents to a secure, authenticated endpoint with a single click. Call via REST or wrap in an MCP server for maximum compatibility with other applications and AI systems.





Aura Agent Resources
Frequently Asked Questions
Charges only accrue for agents made accessible through a public endpoint at a rate of $0.35 per agent/hour. Internal agents are not charged.
Neo4j Aura Agent is available on AuraDB Free, Professional & Business Critical tiers.
Yes, an AuraDB graph database instance is required to build an agent. The agent tools retrieve data from that database.
Google Gemini 2.5 Flash is used for the agent runtime. A fine-tuned version of Gemini Flash is also used for the text2Cypher generation tool. Aura Agent supports various text embedding models from Google VertexAI and Microsoft Azure OpenAI including gemini-embedding-001, text-embedding-005, text-multi-lingual-embedding-002, and OpenAI text-embedding-3-small/large/ada-002.







