# Eigent

> Open-source desktop app that runs a team of AI agents working in parallel

- Category: Agents & Automation / Assistants & Chatbots
- Language: TypeScript
- Licence: Apache-2.0
- Repository: https://github.com/eigent-ai/eigent
- Website: https://www.eigent.ai
- Canonical page: https://ai-tldr.dev/tools/eigent/

## In short

Eigent is an open-source Cowork desktop application that lets you build, manage, and deploy a custom AI workforce. Instead of one chatbot, it runs a team of specialized agents that split a complex task into parts and work on them in parallel. It sits in the Assistants & Chatbots category of the Agents & Automation tools and is open-source, written in TypeScript and licensed Apache-2.0.

## Overview

Eigent is an open-source Cowork desktop application that lets you build, manage, and deploy a custom AI workforce. Instead of one chatbot, it runs a team of specialized agents that split a complex task into parts and work on them in parallel.

It is built on CAMEL-AI's open-source project and ships 100% open source from day one. You can run it fully standalone on your own machine, plug in your own models, and keep complete control over your data with no cloud account required.

Out of the box Eigent includes pre-defined agent workers (Developer, Browser, Document, and Multi-Modal) and a large set of built-in Model Context Protocol (MCP) tools, so agents can browse the web, run code, and connect to apps like Notion, Slack, and the Google suite.

## What it does

- Multi-Agent Workforce that breaks tasks into subtasks and runs several specialized agents in parallel
- Pre-defined agents: Developer (writes and runs code), Browser (searches and extracts web content), Document, and Multi-Modal (images and audio)
- Comprehensive model support, including local models via vLLM, Ollama, and LM Studio
- Built-in MCP tool integration for web browsing, code execution, Notion, Slack, and Google suite, plus the ability to install your own tools
- Human-in-the-Loop: Eigent asks for your input automatically when a task gets stuck or hits uncertainty
- Local deployment with full data control and no required cloud dependency, alongside optional cloud and enterprise (SSO, access control) modes

## Getting started

Local deployment is the recommended way to run Eigent fully standalone. For a fast preview, the cloud-connected Quick Start needs Node.js (version 18-22) and npm, and connects to Eigent's cloud backend (account registration required).

1. Clone, install, and run

   Clone the repository, install dependencies, and start the desktop app in development mode.

```bash
git clone https://github.com/eigent-ai/eigent.git
cd eigent
npm install
npm run dev
```

2. Update dependencies after pulling new code

   After git pull, refresh both the frontend and the Python backend dependencies.

```bash
# 1. Update frontend dependencies (in project root)
npm install

# 2. Update backend/Python dependencies (in backend directory)
cd backend
uv sync
```

3. Prefer fully local

   For complete isolation from cloud services, follow the Local Deployment Guide in the repo. It sets up a local backend server with the full API and local model integration (vLLM, Ollama, LM Studio) with zero external dependencies.

## When to use it

- Plan a detailed travel itinerary with flights, hotels, and activities, then post an HTML report and summary to a Slack channel
- Turn a CSV of bank transactions into a quarterly financial report with charts as an HTML file
- Run market research, such as a UK healthcare overview or a German market-entry feasibility study, and deliver a structured HTML report
- Scan a local folder to find exact and near-duplicate files, or run an SEO audit on a website and return actionable recommendations
