#1010th of 30 in RAG and search
zvec-grep
Local CLI combining ripgrep, BM25 and vector search over code and documents
- Stars
- 4.4k
- License
- Apache-2.0
- Last commit
- Nov 2026
- Last release
- Nov 2026
- Language
- Rust
3 on 9 Jul 20264,368 on 19 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
zg indexes a project into a .zvec-grep/ directory and searches it with exact text or regex, BM25 full-text, and vector retrieval, using the zvec engine. It can be run from the terminal or installed as a search tool for agents such as OpenCode (`zg --install --target opencode`). Results are ranked and linked to source locations; code is indexed with symbols and signatures, prose as sections and chunks.
Who it is for: Developers wanting local semantic and keyword repo search, by hand or via agents
Strengths
- Hybrid, FTS and vector queries in one CLI, with repeatable flags for multiple queries
- Local embedding model option (local/potion-retrieval-32m); files and indexes stay on the machine
- Same index serves terminal use and agent integration on macOS, Linux and Windows
- Benchmarks with paired A/B runs and reproduction docs are published in the repo
Weaknesses
- Agent benefit varies by model and run, per the README's own caveat
- Remote embeddings send data off-machine when enabled
- Index lives in .zvec-grep/ per project and must be built before searching
- Agent integration documented only for OpenCode in the README
What it needs
- Rust, local/potion-retrieval-32m, Qwen3.7 Text Embedding
- Needs Node.js/npm for install, zvec
What changed
From the release notes. No flag means the note does not say so, not that the upgrade is safe.
Also in RAG and search
See all 30| Rank | Project | Score |
|---|---|---|
| 1 | xerjElasticsearch-compatible search server that auto-indexes folders for coding agents | 88 out of 100 |
| 2 | sonicLightweight schema-less search index that returns object IDs, written in Rust | 83 out of 100 |
| 3 | manticoresearchC++ search database with full-text, vector and hybrid search over SQL and HTTP | 83 out of 100 |
| #4 | xbergRust document extraction engine with OCR, tables and 15 language bindings | 82 out of 100 |
| #5 | azure-search-openai-demoRAG chat app over your documents using Azure OpenAI and Azure AI Search | 79 out of 100 |