Harbor: Effortless Local LLM Stack Management with CLI and App

Harbor: Effortless Local LLM Stack Management with CLI and App

Summary

Harbor is a comprehensive containerized toolkit designed to simplify the setup and management of local LLM environments. It allows users to effortlessly run LLM backends, APIs, frontends, and various related services with simple commands or via a companion application. This project streamlines experimentation and development with large language models on your local machine.

Repository Info

Updated on October 12, 2025
View on GitHub

Tags

aillmlocaldockerclipythonself-hostedllm-ops

Click on any tag to explore related repositories

Introduction

Harbor is an innovative, containerized LLM toolkit that empowers users to effortlessly set up and manage their local Large Language Model (LLM) environments. It provides a powerful command-line interface (CLI) and a companion desktop application, enabling the seamless deployment of LLM backends, frontends, APIs, and a wide array of related services. Designed for both beginners and experienced developers, Harbor simplifies the complex process of running cutting-edge AI models and tools locally.

Installation

Getting started with Harbor is straightforward. The project provides comprehensive guides for installing both the CLI and the companion application. Once installed, you can launch a default stack, including Open WebUI and Ollama, with a single command:

harbor up

For more detailed installation instructions, refer to the official documentation.

Examples

Harbor's versatility shines through its wide range of supported services and functionalities. Here are a few examples of what you can achieve:

Running Local LLMs and Inference Engines

Easily deploy popular LLM backends and advanced inference engines, all pre-connected to frontends like Open WebUI:

harbor up ollama
harbor up llamacpp
harbor up vllm
harbor up vllm llamacpp tgi litellm tabbyapi aphrodite sglang ktransformers mistralrs airllm

Web RAG and Deep Research

Integrate powerful search capabilities with your LLMs. Harbor includes SearXNG, which is pre-connected to various services for Web RAG:

harbor up searxng
harbor up searxng chatui

Image Generation

Utilize image generation models like FLUX with ComfyUI, integrated directly into Open WebUI:

harbor up comfyui

Access from Anywhere

Access your Harbor services from your phone via QR code or expose them to the internet with a built-in tunneling service (use with caution):

harbor qr
harbor tunnel

Eject to Docker Compose

When you're ready to transition to a custom setup, Harbor can generate a docker-compose file replicating your current configuration:

harbor eject searxng llamacpp > docker-compose.harbor.yml

Why Use Harbor

Harbor offers significant advantages for anyone working with local LLM environments:

  • Simplicity and Convenience: Run complex LLM stacks with minimal configuration, often with a single command or click.
  • Centralized Workflow: All services, configurations, logs, and data files are managed from a single, consistent interface, reducing setup complexity and ensuring predictability.
  • Experimentation Hub: It serves as an excellent starting point for experimenting with a vast array of LLMs and related services, providing a robust and ready-to-use environment.
  • Extensive Service Catalog: Access a wide selection of UIs, backends, and satellite services, from chat interfaces to RAG tools and workflow automation.

Links