Show HN: Vecdb – local-first hybrid vector database in Rust (HNSW and BM25)
Vecdb is an open-source vector database written in Rust that allows for local-first data storage. It combines dense nearest-neighbor retrieval with sparse keyword search to enhance search precision and recall. Unlike cloud-based alternatives, vecdb operates entirely on local hardware, ensuring privacy and eliminating external dependencies.
- ▪Vecdb supports hybrid semantic search by fusing dense HNSW retrieval with sparse BM25 keyword search.
- ▪The database is self-hosted, requiring no external services or API keys, and is designed to run on personal hardware.
- ▪It is written in pure Rust, ensuring memory safety and high performance, capable of serving thousands of queries per second.
Opening excerpt (first ~120 words) tap to expand
vecdb Open source, production-grade vector database written in Rust. What is vecdb? vecdb is a self-hosted vector database that stores high-dimensional float32 vectors alongside JSON metadata and supports hybrid semantic search — combining dense nearest-neighbour retrieval (HNSW/IVF) with sparse BM25 keyword search. Results from both indexes are fused using a configurable alpha weight so you get the precision of keyword matching and the recall of embedding search in a single query. vecdb differs from managed alternatives in one key way: it runs entirely on your own hardware with no external services, no API keys, no egress fees, and no cloud dependency. It ships as a single statically-linked binary (~10 MB). A fresh instance is running in under a second.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at GitHub.