LibreAI / Dev Tools / grepai
// devtools · yoanbernabeu/grepai

grepai

grepai is an open-source dev tools project by yoanbernabeu (yoanbernabeu/grepai on GitHub). It has 1,902 GitHub stars and 161 forks, is released under the MIT license, and was last updated on GitHub on 21 September 2026.

grepai: semantic code search and call graphs for AI agents, running 100% locally on your machine.

★ 1.9k stars⑂ 161 forksMITupdated 2026-09
▦ visual — this project's README ships no embeddable screenshot,
so here's a placeholder. The guide below is complete regardless.
// project facts

Project facts

GitHub stars
1,902
Forks
161
License
MIT
Last updated on GitHub
21 September 2026
Category
Dev Tools
Listing
Full install guide
Official site
Official site
// how it works

How it works

  1. Clarify your goal

    Decide which embedding provider to use; a local Ollama model keeps everything on your machine.

  2. Get the project

    Install grepai with Homebrew or the install script, and pull the embedding model if you use Ollama.

  3. Install and configure

    Initialise grepai in your project and start the watch daemon to build and maintain the index.

  4. Run your first task

    Search semantically in plain language and trace callers before changing a function; optionally connect it to your agent as an MCP tool.

// features

Why people love it

Search by intent

Natural-language queries find conceptually related code even when names differ, for example finding session handling when you search for authentication logic.

Call graph tracing

Trace who calls a function before changing it, reducing breakage from blind edits.

100% local operation

With local embedding providers, your code never leaves your machine, suiting private and regulated codebases.

Always-fresh index

A file watcher keeps the index up to date automatically as you edit.

Agent-ready and MCP server

Designed to feed coding agents relevant context instead of raw grep output, and exposes itself as an MCP tool agents can call directly.

Shell completion

Generated completions for zsh, bash, fish, and PowerShell cover commands, flags, and dynamic values like workspace and provider names.

Advertisement
// run it

Run it yourself

  1. Clone the project (optional)

    Most users install the released binary instead of building from source.

    git clone https://github.com/yoanbernabeu/grepai.git
    cd grepai
  2. Install and prepare embeddings

    Install via Homebrew (or the README's install script) and pull the default embedding model.

    brew install yoanbernabeu/tap/grepai
    ollama pull nomic-embed-text
  3. Initialise, index, and search

    Run these inside your project.

    grepai init
    grepai watch
    grepai search "error handling"
// use cases

What you can build

  • Letting Claude Code find the auth flow in an unfamiliar monorepo without reading hundreds of files
  • Checking every caller of a function before refactoring it
  • Searching a legacy codebase where naming conventions are inconsistent
  • Giving an agent precise context while keeping proprietary code fully offline
Advertisement
// faq

Questions, answered

Who maintains it?

It is created and maintained by Yoan Bernabeu.

Is it free to use?

Yes, grepai is MIT-licensed and free. With Ollama or LM Studio embeddings there is no per-query cost at all.

Does it need a paid API key?

No paid key is required: Ollama is the recommended default and runs locally. OpenAI embeddings are an optional alternative that would use your key.

How does it compare to a well-known alternative?

ripgrep matches exact text and regex; grepai matches meaning through embeddings and adds call-graph tracing, at the cost of maintaining a local index.

How do I get started?

Install it, run ollama pull nomic-embed-text, then grepai init and grepai watch in your project, and search in plain language.

// keep exploring

Similar tools

Dev Tools

SocratiCode

Enterprise-grade codebase intelligence for AI agents — hybrid semantic search over 40M+ LOC, locally.

★ 3.3k starsAGPL-3.0
View guide →
Dev Tools

semble

Fast, accurate semantic code search for agents — about 99% fewer tokens than grep plus read.

★ 6.2k starsMIT
View guide →
Dev Tools

cocoindex-code

Lightweight AST-based semantic code search CLI (ccc) that makes coding agents faster and cheaper

★ 2.8k starsApache-2.0
View guide →
Dev Tools

probe

Semantic code search for AI — ripgrep speed with tree-sitter AST parsing.

★ 722 starsApache-2.0
View guide →
Dev Tools

accessibility-agents

Eleven AI agents that enforce WCAG 2.2 AA accessibility in your code.

★ 420 starsMIT
View guide →
Dev Tools

ygrep

Fast, local, indexed code search built for AI assistants.

★ 60 starsMIT
View guide →
// alternatives

Alternatives in Dev Tools

The most-starred other Dev Tools projects in the LibreAI catalogue — the realistic shortlist if grepai is not the right fit.

  • OpenCode★ 211.6k

    The open source coding agent

  • Transformers★ 166.9k

    The model-definition framework behind modern AI — thousands of pretrained transformers for text, vision, and audio.

  • claude-code★ 149.7k

    Anthropic's agentic coding tool that lives in your terminal and works across your whole codebase

  • cc-switch★ 141.3k

    Cross-platform desktop all-in-one manager for Claude Code, Codex, OpenCode and other coding agents

  • codex★ 128.2k

    OpenAI's lightweight coding agent that runs locally in your terminal

// more in devtools

More in Dev Tools

Nearby Dev Tools projects in the catalogue, closest in stars to grepai — keep exploring the shelf either side of it.

  • tmuxai★ 2.0k

    AI-Powered, Non-Intrusive Terminal Assistant

  • sourcery★ 1.9k

    Instant AI code reviews

  • memorax-code★ 2.1k

    Shared memory for coding agents: lessons, repo knowledge, and your preferences, carried across sessions.

  • pi_agent_rust★ 1.9k

    A native Rust port of the Pi coding agent: single binary, zero unsafe code, and hardened extensions.

  • claude-code.nvim★ 2.1k

    Seamless integration between Claude Code AI assistant and Neovim

  • repo2txt★ 1.8k

    Turn any GitHub repo into one formatted text file — perfect context for AI.

Get grepai

Free and open — grab the source, star the repo, and start creating.