K-AI

Install K-AI

Installation guides are coming soon.

coming soon

Not published yet. We are writing the install guides now. This page will cover each deployment shape below, with copy-and-paste commands and a short checklist for the first request.

What the guides will cover

Single machine soon

The router and control panel with one local GPU node, using Docker Compose.

Kubernetes soon

The router, control panel, knowledge API and database behind your own single sign-on and ingress.

Adding nodes soon

The node agent on Linux hosts and on Windows with Docker Desktop, including how model files get to the node.

RunPod soon

Connecting your RunPod account, creating a network volume, and tuning idle stop and the spend cap.

Knowledge bases soon

Starting the knowledge API, choosing embedding and extraction models, and ingesting your first documents.

In the meantime, the documentation explains the components and how they fit together.