Your AI Homestead

Discover and serve local AI models with confidence. One CLI to rule them all.

$ lai discover

Scanning Ollama... Done (12 models)

Scanning HF Hub... Done (8 models)

Scanning GGUF files... Done (3 models)

Registry updated. 23 models tracked.

Universal Discovery

Find every model on your machine, regardless of the engine. Ollama, HuggingFace Hub, GGUF files, and MLX—all in one registry.

Instant Serving

Start any discovered model with a single command. Homestead automatically selects the best engine for your hardware.

Deep Observability

Monitor prompt usage, token counts, and performance metrics locally. Keep your data private while getting world-class insights.

Simple by Design

Step 1

Install Homestead

Get the `lai` binary installed on your system with a single script.

# Install Homestead
curl -sSL get.homestead.ai | sh

Step 2

Discover Models

Scan your machine for LLMs. This updates the local registry with metadata and tags.

# Full machine scan
lai discover

Step 3

Serve & Chat

Launch an OpenAI-compatible API server for any model in your registry.

# Start serving
lai serve llama3