nyas
A lightweight vector database for similarity search
// Initialize the vector store let store = VectorStore::new(config); // Add embeddings store.add(vectors, metadata); // Search for similar vectors let results = store.search(query, k: 10);
Built for Performance
Multiple Distance Metrics
Support for L2, Cosine, and Dot Product similarity measures for flexible vector comparison.
Hybrid Execution
Intelligent CPU/GPU hybrid execution for optimal performance across different workloads.
Graph-Based Search
Inspired by DiskANN and FreshDiskANN for fast, accurate billion-point nearest neighbor search.
Flexible Storage
Multiple storage backend options to match your infrastructure and performance needs.
Written in Rust
Memory-safe and blazingly fast implementation with zero-cost abstractions.
Apache 2.0 License
Open source and free to use in commercial projects with a permissive license.