harbor

Stop configuring your AI stack. Start using it. One command brings a complete pre-wired LLM stack with hundreds of services to explore.

What is harbor?

Stop configuring your AI stack. Start using it. One command brings a complete pre-wired LLM stack with hundreds of services to explore.

Overview of harbor

What it is

Harbor is a CLI and companion application that sets up a complete local LLM stack with a single command. It orchestrates Docker Compose to run backends like Ollama, llama.cpp, or vLLM, frontends like Open WebUI, and supporting services such as SearXNG for web search, Speaches for voice chat, and ComfyUI for image generation. All services are pre-configured to work together, eliminating manual setup. Harbor is open source under the Apache-2.0 license and is categorized under AI tools. Its repository is hosted at https://github.com/av/harbor, and community support is available via Discord at https://discord.gg/8nDRphrhSF.

Key features

  • Single-command deployment of a pre-wired LLM stack using harbor up.
  • Supports multiple inference backends: Ollama, llama.cpp, vLLM, Docker Model Runner, MLX, oMLX, TGI, LiteLLM, TabbyAPI, Aphrodite, SGLang, ktransformers, Mistral.rs, AirLLM.
  • Includes frontends like Open WebUI, ChatUI, Morphic, Perplexica, and Local Deep Research, all pre-connected to backends and search services.
  • Provides agentic coding workflows via Harbor Boost, with modules for web research, deliverable audits, and scope checks.
  • Offers built-in tunneling (harbor tunnel) and Traefik reverse proxy for exposing services to the internet.
  • Manages MCP (Model Context Protocol) tools through MetaMCP and mcpo, with a Web UI for configuration.
  • Supports image generation with ComfyUI and Flux, voice chat with Speaches, and LLM workflow builders like Dify, n8n, and LangFlow.

Use cases

  • Running local LLMs for development, testing, or privacy-sensitive applications without cloud dependencies.
  • Setting up a local web RAG (Retrieval-Augmented Generation) system with SearXNG and Open WebUI for deep research.
  • Building and testing agentic coding workflows with Boost, including multi-hop research and deliverable auditing.
  • Experimenting with different inference engines and frontends in a pre-configured environment.
  • Exposing local AI services to a local network or the internet for remote access or mobile use.

Getting started

bash# Starts fully configured Open WebUI and llama.cppharbor up# Now, Open WebUI can do Web RAG and TTS/STTharbor up searxng speachesFor installation guides and a full service catalog, see the repository README.

License and pricing

Harbor is licensed under Apache-2.0 and is open source. No pricing information is stated.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/avharbor.svg)](https://yunodb.com/tools/avharbor?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/avharbor?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/avharbor.svg" alt="Featured on YunoDB" height="20"></a>