Zvec
zvec.org
- Free plan
- Windows
- macOS
- Linux
- iPhone and iPad
- Android
- API
- Self-hosted
What Zvec is
Zvec is a free, open-source vector database that runs inside an application process, without a separate server, daemon, or external infrastructure. It indexes and queries dense and sparse vectors, supports full-text search, and can combine retrieval with structured filters. Listed index choices include Flat, HNSW, IVF, DiskANN, and RaBitQ variants; quantization options include FP16, INT8, INT4, RaBitQ, and PQ. Write-ahead logging preserves data across process crashes or power failures. Multiple processes may read the same collection at once, but only one process can write to it. SDKs are listed for Python, Node.js, Go, Rust, Dart/Flutter, C, and C++. Supported environments include Linux, macOS, and Windows, with prebuilt SDK binaries also available for Android and iOS. Named integrations include Logseq, MyContext, ReMe, txtai, and CocoIndex. The product is positioned for application developers, from local development and prototyping to embedded, edge, and production systems. Zvec is self-hosted and licensed under Apache-2.0.
Who it suits
Zvec suits application developers who want local vector indexing and retrieval without separate infrastructure. Its stated use cases range from prototyping and local development to embedded, edge, and production systems.
Good to know, and what to watch
- Runs locally without a server or external infrastructure
- Supports dense and sparse vectors, plus full-text search
- Offers hybrid retrieval with structured filters
- Write-ahead logging preserves data after crashes or power failures
- SDKs span seven listed languages
- Writes are exclusive to one process
- No free trial is listed
Plans
Every plan and priceZvec
Open source
The verdict
Zvec offers a broad search and indexing toolkit in an in-process, self-hosted package. Its single-process write limit is worth considering for applications that need concurrent writers.
What the maker publishes
- Agent tools
- The product family includes Zvec-Grep, a local workspace search CLI and MCP tool for people and AI agents. source
- Concurrency limit
- Multiple processes can read the same collection simultaneously, while writes are exclusive to one process. source
- Indexes
- The site lists Flat, HNSW, IVF, and DiskANN index types, plus FP16, INT8, INT4, RaBitQ, and PQ quantization. source
- Integrations
- The product site names integrations with Logseq, MyContext, ReMe, txtai, and CocoIndex. source
- Languages and APIs
- Official SDKs are listed for Python, Node.js, Go, Rust, and Dart/Flutter, with C and C++ also linked from the product site. source
- Privacy and deployment
- Zvec runs in-process and locally, with no server or external infrastructure required. source
- Search types
- It supports dense and sparse vectors, full-text search, and hybrid retrieval with structured filters. source
- Security and compliance
- The maker describes Zvec as private and the repository identifies Apache-2.0 licensing; the pages opened do not state a security certification or compliance standard. source
- Storage
- Write-ahead logging provides persistence across process crashes or power failures. source
- Support
- The documentation directs users to report issues on GitHub and lists community channels including DingTalk, WeChat, Discord, and X. source
- Supported content
- Zvec-Grep searches local code, documentation, and text-based data using exact text, keywords, or meaning. source
- Supported platforms
- The README lists Linux, macOS, and Windows, and says prebuilt SDK binaries are published for Linux, macOS, Windows, Android, and iOS. source
- What it does
- Zvec is an open-source, in-process vector database for indexing and querying vectors without a server, daemon, or external infrastructure. source
- Who it is for
- The maker describes Zvec for application developers and presents use cases spanning prototyping, local development, embedded and edge applications, and production systems. source
Checked against the maker's pages on 8 Oct 2026.