Eigent

Eigent is an open-source desktop application built by EIGENT UK LTD that lets individuals and teams deploy a custom AI workforce directly on their machine. It is built on top of CAMEL-AI's multi-agent framework and supports both single-agent and multi-agent (workforce) modes for automating complex, multi-step workflows. The core platform is licensed under Apache 2.0 and available on GitHub, with a cloud-connected managed option also available.

What It Is

Eigent positions itself as a "cowork desktop" — a local-first AI agent environment where users describe tasks in natural language and one or more specialized agents execute them end to end. The docs describe it as "the first open-source, general-agent desktop app" for building, managing, and deploying a custom AI workforce. It supports browser automation, terminal execution, file operations, multi-modal tasks, and scheduled workflows, all running on the user's own hardware.

Multi-Agent Workforce Architecture

The core differentiator is Eigent's workforce model, which dynamically breaks down tasks and activates multiple specialized agents to work in parallel:

• Browser Agent — searches the web, reads pages, and interacts with web UIs
• Terminal Agent — executes shell commands and scripts
• Multi-modal Agent — handles image understanding and visual tasks
• Document Agent — reads, writes, and summarizes files

A single-agent mode is also available for focused, direct tasks. The workforce coordinator assigns subtasks, tracks progress, and surfaces an execution context showing which skills, MCP tools, and files were used.

Open-Source Deployment Model

Eigent is 100% open source under the Apache License 2.0. The recommended path is local deployment: users clone the repository, run a local FastAPI backend (managed with uv), and connect their own model providers — including vLLM, Ollama, and LM Studio — so no data leaves the machine. A cloud-connected quick-start mode is also available for users who prefer managed infrastructure. The GitHub repository had over 14,700 stars and 1,742 forks as of the latest metadata.

Model Agnostic and MCP Integration

Eigent does not lock users into a single model provider. It connects to cloud APIs, enterprise model gateways, or locally hosted inference servers. The docs note over 200 built-in Model Context Protocol (MCP) tools covering web browsing, code execution, and more, with support for installing custom tools or integrating internal APIs. The skill system lets agents dynamically acquire the right tools for a given task or accept manually curated skill sets for governance.

Automation and Scheduling

Beyond interactive cowork, Eigent supports scheduled triggers for recurring workflows — daily standups, weekly reports, SEO briefs, market research — that run autonomously on a defined cadence. Webhook and event triggers are also on the roadmap for real-time external system integration. Human-in-the-loop checkpoints are built in: if an agent gets stuck or encounters uncertainty, it requests human input before proceeding.

Update: v1.0.2 Release

The GitHub repository shows the latest release is v1.0.2, published in July 2026, with the repository created in July 2025. The homepage banner announces "Eigent 1.0 New Version Released." The about page states the product has been live for less than three months at time of writing, with the company reporting over $250K in revenue since launch. The roadmap lists active development areas including context engineering, multi-modal enhancement, browser toolkit improvements, and reinforcement learning environment integration via VERL, TRL, and OpenRLHF. The project is built on CAMEL-AI, which the about page describes as the world's first multi-agent LLM framework, with over 35K GitHub stars across open-source projects.