HelixDB

Loading visualization...

Backed by Y Combinator

Backed by NVIDIA

A Graph-Vector
Database
Built For

Feed your agents more relevant context with the fastest and most scalable graph-vector database, built in Rust.

Star on GitHubTakes 2 seconds 🙏

Getting Started

Install Now

Install

Terminal

> $ curl -sSL "https://install.helix-db.com" | bash
$ helix install

Build

queries.hx

> $ helix init --path

Deploy

POST localhost:6969

> curl -X POST http://localhost:6969 -H

Install HelixDB

Get started with HelixDB in minutes, install the CLI and then initialize a project.

Terminal

> $ curl -sSL "https://install.helix-db.com" | bash

This will install the Helix CLI to your machine.

Terminal

> mkdir helix-project && cd helix-project

Initialize the project

> helix init

Initialize Project

Now that the Helix CLI is installed on your machine, initialize a project.

Hybrid Query Traversals

Seamlessly combine vector similarity search with graph traversals in a single, powerful query. No more complex joins or multiple database calls.

QUERY findSimilarFriends(userID: String, queryVec: Vector) =>
similar <- SearchV(queryVec, topK: 5)
friends <- similar::Out<Friends>
RETURN friends::{ ID, name, similarityScore }

Vector Entry

Connected Node

Type-Safety

Advanced static analysis provides real-time feedback, autocomplete, and error detection. Write queries with confidence.

Type Checker

> helix check
❌ Checking Helix queries
error: 'Know' is not a valid edge type (in QUERY named 'get_user*)
      |--queries.hx: 16:38
16    |   user_nodes <- N<User> (node_1d):: 0ut<Know>
---> help: check the schema for valid edge types
...

Deploy

> helix deploy --local --path helixdb-cfg
✓ Compiling Helix queriesNodes(Some("User"))
Successfully compiled 2 query files
Successfully transpiled queries
Successfully wrote queries file
Successfully built Helix
Successfully started Helix instance
└── Instance ID: aafd3dc7-21cb-4073-b330-aec71e48b623
└── Port: 6969
└── Available endpoints:
    └── /get_user

Lower Costs

Eliminate the complexity and cost of maintaining separate vector and graph databases. One unified solution.

Database

Cloud

50%

Less Data

High Speeds

Optimized for both vector similarity and graph traversal workloads with industry-leading performance metrics.

~2ms

Vector Similarity Search

Sub 1ms

Graph Traversals

Join Our Growing Community

Be part of the next generation of database technology. Connect with developers and innovators building the future.

Join Discord Star on GitHub

GitHub

Discord

Active

Uptime99.99%

Response Time<10ms

Auto-scalingEnabled

Coming Soon

Managed Cloud Service

Automatic scaling to handle traffic spikes
24/7 expert support and monitoring
Enterprise-grade security and compliance

Join Waitlist

Be the first to know when we launch

You've made it this far!

If you're excited about what we're building, show your support with a star.
It helps us grow and motivates us to keep pushing the boundaries.

Star on GitHub

⚡Takes 2 seconds🚀Join our developers

A Graph-Vector Database Built For

Getting Started

Hybrid Query Traversals

Type-Safety

Lower Costs

High Speeds

Join Our Growing Community

Managed Cloud Service

You've made it this far!

A Graph-Vector
Database
Built For