HelixDB is an open-source database designed to unify graph and vector data models into a single high-performance system.

Built in Rust, it enables developers to store, query, and traverse both structured relationships (graphs) and semantic embeddings (vectors) within one database, removing the need to combine separate systems for graph databases, vector databases, and application-layer orchestration.

It is primarily designed for AI-native applications such as retrieval-augmented generation (RAG), agent memory systems, and semantic search workflows. HelixDB introduces its own type-safe query language (HelixQL) and supports features like compiled queries, embeddings integration, and efficient graph traversal. Its architecture emphasizes low-latency performance using an optimized storage engine (LMDB-based) and compiled execution paths for queries.

The system also includes SDKs for Python, TypeScript, and other languages, along with a CLI for local development and cloud deployment options. It is positioned as a foundational infrastructure layer for building intelligent applications that require both relational reasoning and semantic similarity search.

Key features include:

• Unified graph + vector database model
• Rust-based high-performance engine
• Type-safe query language (HelixQL)
• Built-in embedding support for AI workloads
• Low-latency graph traversal and vector search
• Python and TypeScript SDKs
• Local and cloud deployment options
• Designed for RAG and AI agent memory systems

Common use cases include:

• Building AI agents with long-term memory
• Powering retrieval-augmented generation systems
• Semantic search over connected data
• Knowledge graphs with embeddings
• Backend infrastructure for AI-native applications