zvec-grep

zvec-grep is a local-first search layer that unifies ripgrep, BM25, and vector search for humans and agents.

zg (zvec-grep) is a local-first search layer built for humans and AI agents. It unifies ripgrep, BM25, and vector search behind a single interface. Users and agents can run it directly from the terminal or connect it to tools like OpenCode, Codex, Claude Code, and Cursor. Files, indexes, and local models remain on the user machine while remote embeddings require explicit permission. It supports searching source code, documents, and structured data on macOS, Linux, and Windows.

zvec-grep at a glance

Best for: Software engineer
Runs on: macOS, Linux, Windows, CLI
Pricing: Free
Released: July 2026
Category: AI search engine
Website: github.com/zvec-ai/zvec-grep

Key features

Unify ripgrep, BM25, and vector search
Run searches from the terminal or via agents
Index source code, documents, and structured data
Keep files, indexes, and local models on your machine
Configure embedding concurrency for local and remote models
Diagnose indexing errors with debug options

Use it for

Search local workspaces using semantic and lexical queries
Allow AI agents to retrieve code and document evidence
Analyze codebases with symbol-aware retrieval

Alternatives to zvec-grep

jevgrep — Reduces coding agent costs by providing targeted, relevant repository context.
Sgrep — Runs locally without accounts or hosted indexes using small CPU models.
XERJ — It answers agent questions from an index instead of wasting tokens on context windows.
TurboVec — Fits a 10 million document corpus in 4 GB of RAM and searches faster than FAISS.
Anygrep — Runs locally on the CPU using EmbeddingGemma 2 without sending data away.
Gsearch — Runs a real browser over MCP to fetch genuine Google results with zero API keys.
SearchCLI — Provides a stable, reviewable CLI for managing Volcengine AI search workflows.
BGE Embeddings — Provides specialized models for high-performance information retrieval.
SereneDB — Executes search and analytics queries with a PostgreSQL-compatible frontend.
Vespa — Combines vector database capabilities with distributed machine-learned model inference.
Orama — Runs in browsers, servers, and edge networks with zero dependencies.
RuVector — Provides persistent vector retrieval, graph relationships, and local decision making in Rust.

All zvec-grep alternatives