Nornicdb
Nornicdb is a distributed low-latency, Graph+Vector, Temporal MVCC with all sub-ms HNSW search, graph traversal, and writes. Using Neo4j Bolt/Cypher and qdrant's gRPC means you can switch with no changes. Then, adding intelligent features like schema
What is Nornicdb?
Nornicdb is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to nornicdb is a distributed low-latency, graph+vector, temporal mvcc with all sub-ms hnsw search, graph traversal, and writes. using neo4j bolt/cypher and qdrant's grpc means you can switch with no chan...
Nornicdb is a distributed low-latency, Graph+Vector, Temporal MVCC with all sub-ms HNSW search, graph traversal, and writes. Using Neo4j Bolt/Cypher and qdrant's gRPC means you can switch with no changes. Then, adding intelligent features like schema
This server falls under the Databases category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Nornicdb is a distributed low-latency, Graph+Vector, Tempora
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx nornicdbConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Nornicdb
NornicDB is a distributed Graph+Vector database that delivers sub-millisecond HNSW vector search, graph traversal via Neo4j-compatible Cypher/Bolt protocol, and Qdrant-compatible gRPC — all in a single engine with temporal MVCC support. It features on-device embedding with BGE-M3 models, a knowledge decay system modeled on the Ebbinghaus forgetting curve, and over 950 APOC functions for in-database ML and text processing. Teams use it as a drop-in replacement for Neo4j or Qdrant that adds native AI memory management and graph-RAG capabilities without running multiple separate services.
Prerequisites
- Docker installed (recommended for quickest setup) or Go 1.21+ for building from source
- Ports 7474 (HTTP/Admin UI) and 7687 (Bolt) available
- At least 4 GB RAM; 8 GB recommended for embedding workloads
- NVIDIA GPU with CUDA drivers (optional, for GPU-accelerated vector search)
- A Neo4j-compatible client library (e.g., neo4j Python driver) or Cypher-capable MCP client
Start NornicDB with Docker
Pull and run the appropriate Docker image for your hardware. Use the arm64-metal image for Apple Silicon, amd64-cpu for standard x86, or amd64-cuda for NVIDIA GPU acceleration.
# Apple Silicon
docker run -d --name nornicdb -p 7474:7474 -p 7687:7687 \
-v nornicdb-data:/data timothyswt/nornicdb-arm64-metal-bge:latest
# AMD64 CPU
docker run -d --name nornicdb -p 7474:7474 -p 7687:7687 \
-v nornicdb-data:/data timothyswt/nornicdb-amd64-cpu-bge:latestVerify the admin UI is running
Open the browser-based admin interface to confirm NornicDB started successfully and to run your first Cypher queries interactively.
open http://localhost:7474Connect with a Neo4j-compatible client
Use any Neo4j Bolt driver to connect to NornicDB. The Python neo4j driver works without any code changes.
pip install neo4j
python3 -c "
from neo4j import GraphDatabase
driver = GraphDatabase.driver('bolt://localhost:7687')
with driver.session() as s:
s.run(\"CREATE (n:Memory {content: 'Hello NornicDB'})\"
result = s.run('MATCH (n:Memory) RETURN n.content')
print([r['n.content'] for r in result])
"Run a vector similarity search
Use Cypher's vector index query to perform semantic search over embedded content. NornicDB runs embeddings locally using BGE-M3.
CALL db.index.vector.queryNodes('embeddings', 10, 'machine learning')
YIELD node, score
RETURN node.content, score
ORDER BY score DESCConfigure decay profiles for memory management
Create a decay profile to model knowledge forgetting over time using the Ebbinghaus curve. This is useful for AI agent memory systems where older, unused facts should fade.
CREATE DECAY PROFILE working_memory OPTIONS {
halfLifeSeconds: 604800,
function: 'exponential',
visibilityThreshold: 0.10
}Connect Claude Desktop via MCP
Add NornicDB as an MCP server in Claude Desktop so Claude can query your graph and vector data using natural language.
Nornicdb Examples
Client configuration
Configure Claude Desktop to connect to a running NornicDB MCP server.
{
"mcpServers": {
"nornicdb": {
"command": "npx",
"args": ["nornicdb"],
"env": {
"NORNICDB_BOLT_URL": "bolt://localhost:7687",
"NORNICDB_HTTP_URL": "http://localhost:7474"
}
}
}
}Prompts to try
Example prompts for querying graph and vector data in NornicDB through Claude.
- "Find the 5 most semantically similar nodes to 'neural network training'"
- "Show me all relationships between Person nodes in the knowledge graph"
- "Create a node representing our Q3 product roadmap and link it to relevant team members"
- "Query the temporal history of the 'config' node to see all changes in the last 30 days"
- "Which memories have decayed below the visibility threshold this week?"Troubleshooting Nornicdb
Bolt port 7687 is already occupied by a running Neo4j instance
Stop your existing Neo4j service ('neo4j stop' or via Neo4j Desktop) before starting NornicDB, or change NornicDB's bolt_port in nornicdb.yaml to an unused port like 7688 and update your client connection strings accordingly.
Vector search returns no results even though data was inserted
Confirm embedding is enabled in nornicdb.yaml ('embedding.enabled: true') and that the BGE-M3 model file was downloaded. Check the Docker logs with 'docker logs nornicdb' for any model loading errors.
Docker container exits immediately on Apple Silicon
Ensure you are using the arm64-metal image, not the amd64 image. Run 'docker pull timothyswt/nornicdb-arm64-metal-bge:latest' explicitly and check that Docker Desktop is configured to use Apple Virtualization Framework rather than QEMU.
Frequently Asked Questions about Nornicdb
What is Nornicdb?
Nornicdb is a Model Context Protocol (MCP) server that nornicdb is a distributed low-latency, graph+vector, temporal mvcc with all sub-ms hnsw search, graph traversal, and writes. using neo4j bolt/cypher and qdrant's grpc means you can switch with no changes. then, adding intelligent features like schema It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Nornicdb?
Follow the installation instructions on the Nornicdb GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Nornicdb?
Nornicdb works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Nornicdb free to use?
Yes, Nornicdb is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Nornicdb Alternatives — Similar Databases Servers
Looking for alternatives to Nornicdb? Here are other popular databases servers you can use with Claude, Cursor, and VS Code.
Excelize
★ 20.6kGo language library for reading and writing Microsoft Excel™ (XLAM / XLSM / XLSX / XLTM / XLTX) spreadsheets
MCP Toolbox for Databases
★ 15.3kOpen source MCP server specializing in easy, fast, and secure tools for Databases.
DBHub
★ 2.8kA universal database gateway MCP server that enables AI assistants to connect to and query multiple databases (PostgreSQL, MySQL, MariaDB, SQL Server, SQLite) with support for schema exploration, SQL execution, and secure connections via SSH tunnels.
Tabularis
★ 2.1kA lightweight, cross-platform database client for developers. Supports MySQL, PostgreSQL and SQLite. Hackable with plugins. Built for speed, security, and aesthetics.
Postgres AI Guide
★ 1.7kMCP server and Claude plugin for Postgres skills and documentation. Helps AI coding tools generate better PostgreSQL code.
Anyquery
★ 1.7k🏎️ 🏠 ☁️ - Query more than 40 apps with one binary using SQL. It can also connect to your PostgreSQL, MySQL, or SQLite compatible database. Local-first and private by design.
Browse More Databases MCP Servers
Explore all databases servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Nornicdb in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Nornicdb?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.