Summary
| Requirement | Required? | Why |
|---|
| Python 3.10+ | Yes | Installs moorcheh-edge (CLI + SDK) |
| Docker | Yes | Runs the edge server container via moorcheh-edge up |
| Ollama | For answer only | Local LLM for RAG (qwen2.5:0.5b-instruct) |
How install and startup work
pip install moorcheh-edge
moorcheh-edge up
| Step | What it does |
|---|
pip install moorcheh-edge | Installs the moorcheh-edge CLI and Python SDK |
moorcheh-edge up | Pulls moorcheh/moorcheh-edge:latest (if needed), starts Docker container, creates ~/.moorcheh-edge/data |
You do not need Rust or a local source checkout. The server runs from the published Docker image.
Python
Version: Python 3.10 or newer
python --version
pip install moorcheh-edge
Docker
Verify Docker is running:
docker --version
docker info
moorcheh-edge up, Docker pulls moorcheh/moorcheh-edge:latest from Docker Hub and downloads the BGE embedding model (~210 MB, one-time) to ~/.moorcheh-edge/models.
Ollama (answer / RAG)
Required only for moorcheh-edge answer, not for search or upload.
On Linux, moorcheh-edge up --with-llm installs Ollama (into ~/.moorcheh-edge/ollama/) and pulls qwen2.5:0.5b-instruct (~400 MB). Use --skip-ollama on up if you only need vector search.
Ports
| Port | Service |
|---|
| 8080 | Default Moorcheh Edge API |
| 11434 | Ollama (host, when using answer) |
moorcheh-edge up --server-port 8083
moorcheh-edge status --base-url http://localhost:8083
Next steps
