Graph-vector database built in Rust for developers building AI, agent, and knowledge applications.
Website · Docs · Open Source
HelixDB combines graph traversal, vector search, and HelixQL, a strongly typed compiled query language, in one platform. Build locally with the open-source core, deploy with Helix Cloud, and scale into larger production workloads without stitching together separate graph and vector systems.
For teams evaluating larger deployments, Helix Enterprise uses an object-storage-backed architecture for large-scale, multi-tenant search and retrieval of knowledge and context.
- Graph and vector workloads in one system for AI, retrieval, and knowledge applications
- Strongly typed, compiled HelixQL queries for safer developer workflows
- Built in Rust for high performance and low-latency query execution
- Private-by-default deployments with authenticated query access
Helix Enterprise uses Helix's hyperscale architecture to run graph-vector workloads across very large datasets stored in S3-compatible object storage.
- Stateless cluster nodes for flexible horizontal and vertical scaling
- RBAC, monitoring, and enterprise controls
- Supports essentially infinite scaling across TBs of data