AgentIndex icon
AgentIndex
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Submit Tool
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Home/
RAG / Knowledge Base/
zvec
zvec logo

zvec

Active·★ 16.3k·Apache-2.0·Updated 2026-11-17
★ Most Popular

Zvec is an open-source, in-process vector database created and battle-tested within Alibaba Group (Apache 2.0). Designed to operate as an embedded engine within host applications—akin to how SQLite functions for relational storage—Zvec eliminates the operational overhead, network latency, and memory bloat associated with maintaining separate client-server vector infrastructure.

Architecturally, it supports both dense and sparse embeddings, multi-vector queries, and multiple indexing strategies, including memory-resident HNSW and disk-scaling DiskANN with advanced RaBitQ and INT4 quantization. Uniquely for multilingual developer workloads, Zvec natively embeds the Jieba dictionary and HMM tokenization model, enabling frictionless BM25 full-text keyword retrieval and hybrid vector search without requiring external text tokenizer dependencies.

Zvec ships official SDKs across Python, Node.js, Go, Rust, Java, and Dart/Flutter, alongside companion tools like `zvec-grep` (`zg`) for command-line file retrieval. Teams considering Zvec should note its in-process concurrency model: while it supports concurrent multi-process reads, write operations are strictly single-process exclusive, meaning distributed multi-node ingest architectures require an external coordination layer.

This write-up is based on the project's README and public repository data. We have not hand-tested the tool. Data as of 2026-11-12.

License
Apache-2.0
Latest release
v0.8.1 · 2026-11-06
Last commit
2026-11-11
Contributors
40
Forks
1,022
Open issues and PRs
101
First published
2025-12-05
PyPI package
zvec 0.8.1 · 7,641 downloads last week

Source: GitHub, npm and PyPI public APIs.

#agent-skills#db#dev-tooling#embedded#faiss#fts#fulltext-search#hnsw
$ Install
$ pip install zvec
↗ Visit site★ GitHub
01

Features

01Dense, sparse and multi-vector query support
02Full-text search combined with structured filtering
03Embedded local collections with multiple language SDKs
02

Why choose it

+In-process embedded architecture eliminates external server deployments and network latency.
+Hybrid search fuses dense/sparse vector similarity with BM25 keyword search and structured filters.
+Native multilingual support with built-in Jieba Chinese tokenization and HMM models.
+Broad platform and SDK coverage across Python, Node.js, Go, Rust, Java, Flutter, and edge devices.
03

Trade-offs

!Single-writer concurrency design restricts writes to one exclusive process at a time.
!Building large DiskANN indexes and running quantization compression temporarily demands high CPU and memory.
!Native C++ binding dependencies require compatible prebuilt wheels or toolchains on legacy Linux kernels.
!Lacks built-in distributed clustering; horizontal sharding and cross-node replication require manual middleware.
04

Maintenance check

ActiveChecked: 2026-11-12
Last commit
2026-11-11
Commits, last 90 days
100+
Releases, last 90 days
6
Latest release
v0.8.1 (2026-11-06)
Open issues
101
Recent issues with at least one comment
5/7
Commits per week, last 12 weeks
Commits per week, last 12 weeks: 11, 7, 16, 19, 7, 2, 1, 4, 15, 7, 18, 2

Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.

Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.

Most discussed open issues

  • #868 [Performance]:High-selectivity range filter materializes a full bitmap — `bitset_set_list` ~49% CPU, dwarfing the ANN search12 comments · 2026-11-02
  • #586 [Feature]:未来是否会支持单写多读(一个写进程 + 多个只读进程)?4 comments · 2026-07-10
  • #293 clang-tidy: enable bugprone-*4 comments · 2026-03-31
  • #781 Bug: Collection.group_by_query returns empty results with ID-out-of-range errors (UINT64_MAX)3 comments · 2026-09-24
  • #291 clang-tidy: enable performance-*3 comments · 2026-03-31

Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.

05

Compatibility

Python
64-bit Python 3.10–3.14 documented
Verified via docs
Node.js
Official @zvec/zvec SDK documented
Verified via docs
Other SDKs
Go, Rust, Java and Dart/Flutter documented
Verified via docs
06

Quick start

1
$ pip install zvec
07

Use cases

↳Embedding semantic search into an application
↳Combining keyword and vector retrieval for a knowledge base
↳Prototyping local retrieval without running a database server
08

Alternatives

⚡Top VS:
VSbook-to-skill★38.3k→VSOpenKnowledge★4.9k→VSawesome-claude-video-skills★2.3k→VSsynthadoc★2.0k→
book-to-skill logo
book-to-skill★ 38.3k
Turn any technical book PDF into a Claude Code skill — ready to study, reference, and use while you work.
VSCompare→
OpenKnowledge logo
OpenKnowledge★ 4.9k
Beautiful, AI-native markdown IDE and LLM wiki
VSCompare→
nihaixia logo
nihaixia★ 4.1k
倪海厦视角的中医Agent Skill,基于倪海厦教学资料开发,蒸馏倪师伤寒论、金匮要略、黄帝内经、神农本草经、针灸篇等,人纪/医案/经方思维,六经辨证,八纲辨证,天机道,天纪,紫微斗数,易经,阴阳,八卦,五行,风水,地纪等,总结8个诊断公式+快速诊断流程图+脉舌速查+七步走思维模式,蒸馏129条伤寒论 · 23篇金匮 · 72篇黄帝内经 · 神农本草经374种本草(上137/中110/下127) · 1257 例结构化案例 + 243 例叙事医案 · 2,452页讲义。
awesome-claude-video-skills logo
awesome-claude-video-skills★ 2.3k
Open-source skills and toolkits that let Claude Code, Codex and other coding agents make video. 180 repos by type, each security-graded. English / 中文.
VSCompare→
synthadoc logo
synthadoc★ 2.0k
Synthadoc: An open-source LLM knowledge compilation engine that turns raw documents into structured, local-first wikis. A transparent, human-readable alternative to traditional RAG, which can be self-managed and self-improved without the use of any tools.
VSCompare→
awesome-agent-skills logo
awesome-agent-skills★ 37.3k
A curated collection of 1000+ agent skills from official dev teams and the community, compatible with Claude Code, Codex, Gemini CLI, Cursor, and more.
VSCompare→
Anthropic Skills logo
Anthropic Skills★ 196.7k
Public repository for Agent Skills
VSCompare→
taste-skill logo
taste-skill★ 117.7k
Agent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
VSCompare→
See all alternatives →

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Why choose it03Trade-offs04Maintenance check05Compatibility06Quick start07Use cases08Alternatives
alibaba /
zvec
↗

Lightweight in-process vector database with hybrid search, DiskANN, and built-in Chinese tokenization

16.3k1.0kApache-2.0
Last commit1d ago
StatusActive
CategoryRAG / Knowledge Base
Contributors40
Latest releasev0.8.1
Links
Documentation↗Discussion↗Issues↗Releases↗
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
Best by use caseAboutContactAdvertisePrivacy PolicyTerms of UseQYSGitHubBuy me a coffee ☕

Browse by Category

MCP ServersCode AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.