Developers who want a ready made local AI stack without manually wiring together separate services.

Harbor is a CLI and companion app that runs a pre configured set of open source LLM tools such as backends, front ends, and utilities together via Docker Compose.

Runs locally on any Docker capable machine, orchestrating dozens of interchangeable services with one command.

Actively developed since 2024 by maintainer av with regular service additions.

Removes the tedious configuration work of pairing inference backends with UIs, vector stores, and other companion tools.

Depends on Docker and Docker Compose and integrates services like Ollama, llama.cpp, and Open WebUI.

Open source under Apache 2.0 on GitHub, no account required, everything runs locally in containers.