Harbor
Run any LLM backend, frontend, and API with a single command using containers
Visit Product
460 upvotes
4,247 views
About
Harbor is an open-source containerized environment for running any combination of LLM backends, APIs, frontends, and services with minimal configuration. Think of it as Docker Compose optimized specifically for the AI infrastructure stack — it manages the complex dependencies, networking, and configuration between components like Ollama, Open WebUI, LiteLLM, Langfuse, and dozens of other AI tools so users don't have to.
The project solves a real pain point: running a local AI stack with multiple components requires configuring networking between containers, setting up authentication, managing environment variables, and understanding the dependencies of each tool. Harbor abstracts all this complexity into simple CLI commands, making it possible to spin up a complete, production-grade local AI environment in minutes.
Harbor is used by developers experimenting with local AI stacks, organizations building private AI infrastructure, and researchers who need reproducible environments with specific tool combinations. Its service catalog covers the most commonly used components in local AI deployments.
The project solves a real pain point: running a local AI stack with multiple components requires configuring networking between containers, setting up authentication, managing environment variables, and understanding the dependencies of each tool. Harbor abstracts all this complexity into simple CLI commands, making it possible to spin up a complete, production-grade local AI environment in minutes.
Harbor is used by developers experimenting with local AI stacks, organizations building private AI infrastructure, and researchers who need reproducible environments with specific tool combinations. Its service catalog covers the most commonly used components in local AI deployments.
Product Features
- One-command setup for complete AI infrastructure stacks
- 50+ pre-configured AI services and tools
- Ollama, Open WebUI, LiteLLM, Langfuse support
- Automatic networking between services
- Simple CLI for service management
- Configuration profiles for different setups
- GPU passthrough configuration
- Custom service addition with templates
- Environment variable management
- Docker Compose based for full transparency
- 50+ pre-configured AI services and tools
- Ollama, Open WebUI, LiteLLM, Langfuse support
- Automatic networking between services
- Simple CLI for service management
- Configuration profiles for different setups
- GPU passthrough configuration
- Custom service addition with templates
- Environment variable management
- Docker Compose based for full transparency
About the Publisher
Harbor was created as an open-source project to lower the barrier to running complete local AI stacks. The project fills a practical gap between individual tools that are easy to install in isolation and the challenge of running multiple interconnected AI services together. By providing pre-configured, tested service combinations, Harbor saves developers and organizations the considerable time required to figure out integration details for each new tool combination.