About Zvec
Zvec is an open source, in process vector database from Alibaba. It runs as a library inside the application rather than as a separate server, so a single package is installed and searching starts with no servers or configuration. Because it runs in process, it can live wherever the code does, including notebooks, servers, command line tools and edge devices.
It supports both dense and sparse vectors, with native multi vector queries in a single call, and hybrid search that combines semantic similarity with structured filters. The vendor states it scales to billions of vectors with sub millisecond search latency. Durability comes from a write ahead log, so data persists through a process crash or power failure. Multiple processes can read the same collection at once, while writes are exclusive to a single process.
The project is aimed at developers building retrieval augmented generation, semantic search and recommendation features who want an embedded store instead of operating a database service. The site provides documentation, a quickstart, benchmarks, a blog, an API reference and a GitHub repository, with a choice of languages.
Zvec is free. The site's structured data lists the price as zero, and there are no paid plans or hosted tiers published.