mudler/LocalAGI

What is it?

Key features

• 01No-code agent creation and management through a Web UI.
• 02Exposes an OpenAI Responses-compatible API for every agent.
• 03Built-in knowledge base for RAG with file uploads and semantic search.
• 04Connectors for Discord, Slack, Telegram, GitHub Issues, IRC and Email.
• 05MCP client support for external tools, plus an MCP server endpoint for managing agents.
• 06Periodic tasks scheduled with cron-like syntax.

Use cases

• Running a private AI assistant on your own machine without sending data to cloud APIs.
• Building a Slack or Telegram bot backed by a local model.
• Automating recurring tasks such as scheduled reports or issue triage on GitHub.
• Prototyping agent workflows that other tools can drive through the OpenAI-compatible API.

Quick start

1. 1

   Clone the repository

   git clone https://github.com/mudler/LocalAGI
2. 2

   Enter the project directory

   cd LocalAGI
3. 3

   Start the default CPU setup with Docker Compose

   docker compose up
4. 4

   Open the Web UI at http://localhost:8080 to manage your agents

Things to know

• !Running the default models requires Docker and sufficient local hardware; GPU setups need NVIDIA, Intel, or AMD drivers.
• !Setup for the built-in knowledge base and models can take significant time and disk space.

Topics

FAQ

What is LocalAGI?

A self-hostable AI agent platform that runs locally on CPU or GPU, with a no-code Web UI and an OpenAI Responses-compatible API. LocalAGI is a platform for building AI agents, automations and chat bots that run entirely on your own hardware. Agents are configured through a Web UI without writing code, and each agent exposes an API compatible with OpenAI's Responses API, so existing clients can call it. The project targets consumer-grade hardware and works with both CPU and GPU setups.

Is LocalAGI free to use? What license does it have?

LocalAGI is open source under the MIT license. Check the license terms before using it in commercial products.

Is LocalAGI still maintained?

Actively maintained. The last commit to LocalAGI was on Oct 16, 2026.

What can I use LocalAGI for?

Running a private AI assistant on your own machine without sending data to cloud APIs. Building a Slack or Telegram bot backed by a local model. Automating recurring tasks such as scheduled reports or issue triage on GitHub. Prototyping agent workflows that other tools can drive through the OpenAI-compatible API.

How do I install LocalAGI?

Clone the repository: git clone https://github.com/mudler/LocalAGI Enter the project directory: cd LocalAGI Start the default CPU setup with Docker Compose: docker compose up Open the Web UI at http://localhost:8080 to manage your agents

Where can I get LocalAGI?

The source code and installation instructions are on GitHub: https://github.com/mudler/LocalAGI