Confirm the deployment goal
Decide whether you need provider flexibility, infrastructure ownership, and deeper operational control.
Self-hosting guide
Understand the Sage self-hosting path before you commit to setup work.
This guide explains the self-hosted product story, who it is for, and when to continue into the deeper repository documentation for implementation detail.
Who should self-host Sage
Self-hosting is usually the right move when your team needs:
- control over provider selection
- ownership of support services and deployment
- stronger privacy and operational guarantees
- direct access to deeper architecture and runtime tuning
How it works
Use this page as the public entry point.
Review the product-facing self-hosted story
Use the website pages to understand the runtime, docs split, and operator experience before implementation detail.
Continue into repo documentation
Move into getting started, configuration, and operations docs in the repository once you are ready to deploy.
Next steps
Keep the next click intentional.
Self-hosted Discord AI page
Start with the search-intent landing page that frames the self-hosted story clearly.
Open pageFull getting started guide
Continue into the repository guide when you need the exact setup commands and environment shape.
Open repo guideConfiguration reference
Use the repository reference when you need the complete environment variable and runtime surface.
Open configuration reference FAQ
Frequently asked questions
What makes Sage self-hosting different from a generic bot deploy?
Sage is positioned as a Discord AI runtime, so self-hosting is about owning the product surface behind memory, research, approvals, and operator workflows rather than only running a bot token somewhere.
Do I have to understand the full architecture before trying Sage?
No. The website self-hosting guide exists so you can understand fit first, then move into the architecture and configuration docs when you need them.
Where are the exact commands and environment variables?
The full getting started and configuration reference remain in the repository, linked from this guide once you are ready for implementation depth.