An embedded graph database for AI applications and graph analytics

IssunDB is an embedded graph database written in Rust. It can be embedded in Rust applications without a separate server, and is designed for applications such as GraphRAG pipelines and knowledge graphs.

You can download the latest binaries (for IssunDB CLI, MCP and HTTP servers) from here.

Key Features

ACID transactions, property graph model, and Cypher query support
Graph traversal and analytics using sparse matrix operations
Vectorized query execution with multi-core parallel processing
Vector, full-text, and hybrid search
APIs for Rust, Python, CLI, HTTP REST, and MCP
Support for Linux, macOS, and Windows

[!IMPORTANT] This project is still in early development, so bugs and breaking changes are expected. Please use the issue page to report bugs or request features.

Quickstart

To use IssunDB in your Rust project, add the dependency to your Cargo.toml:

[dependencies]
issundb = "0.1.0-alpha.15"
serde_json = "1.0"

[!NOTE] IssunDB needs Rust 1.85.0 or newer.

The following example opens a database, inserts nodes, creates relationships, and queries the graph using Cypher:

use std::path::Path;
use issundb::{Graph, GraphQueryExt};

fn main() -> Result<(), Box<dyn std::error::Error>> {
    // Open a graph database (with a 1 GB memory map size limit)
    let graph = Graph::open(Path::new("./issundb-data"), 1)?;

    // Add two nodes with properties
    let alice_props = serde_json::json!({ "name": "Alice", "age": 30 });
    let alice_id = graph.add_node("Person", &alice_props)?;

    let bob_props = serde_json::json!({ "name": "Bob", "age": 28 });
    let bob_id = graph.add_node("Person", &bob_props)?;

    // Add an edge between the nodes
    let edge_props = serde_json::json!({ "since": 2021 });
    graph.add_edge(alice_id, bob_id, "KNOWS", &edge_props)?;

    // Optional: rebuild CSR snapshot manually after bulk writes
    graph.rebuild_csr()?;

    // Run a Cypher query and print the results
    let result = graph.query(
        "MATCH (a:Person)-[r:KNOWS]->(b:Person) RETURN a.name, b.name, r.since"
    )?;

    for record in result.records {
        println!(
            "Match: {} knows {} since {}",
            record.values[0],
            record.values[1],
            record.values[2]
        );
    }

    Ok(())
}

# Output:
Match: "Alice" knows "Bob" since 2021

Running IssunDB in a Container

CLI

# Run IssunDB with the CLI
docker run --rm -it -v issundb-data:/data ghcr.io/issundb/issundb:latest

HTTP (REST)

# Run IssunDB with the HTTP API on port 7474
docker run --rm -p 7474:7474 -v issundb-data:/data ghcr.io/issundb/issundb:latest issundb-rest

MCP

# Run IssunDB with the MCP API on port 8000
docker run --rm -p 8000:8000 -v issundb-data:/data ghcr.io/issundb/issundb:latest issundb-mcp

Documentation

The project documentation is available here. The Rust API documentation is available on docs.rs/issundb.

Rust Examples

Refer to the issundb-examples crate for Rust API examples.

Python Examples

Refer to the issundb-py/examples directory for Python API examples.

Knowledge Graph Demo

Contributing

See CONTRIBUTING.md for details on how to make a contribution.

License

IssunDB is available under either of these licenses:

MIT License (LICENSE-MIT)
Apache License, Version 2.0 (LICENSE-APACHE)

Acknowledgements

The logo is from SVG Repo with some modifications.

Issun DB

README