> ## Documentation Index
> Fetch the complete documentation index at: https://delivery.vexa.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Vexa Delivery

> Your running self-hosted Vexa service that stays up to date.

**Your running self-hosted [Vexa](https://github.com/Vexa-ai/vexa) service that stays up to
date.**

1. **Bring your own cloud** — [k8s](environments/kubernetes), [GCloud](environments/gke), [AWS](environments/eks), [OpenShift](environments/openshift). A namespace in a cluster, or an
   entire fresh one.
2. **Install Vexa Delivery** — [new deployment](install) · [existing one](upgrade).
3. **Set it to consume the public images — or a private channel** — [how channels work](how-it-works).
4. **Cluster running, self-updating on new images published** — [operations](operations).

Everything is open source, images public on Docker Hub. Nothing reaches in: your cluster
pulls, verifies with keys you pin, and promotes to production on your attestation.

Already running Vexa? Start at [Upgrade](upgrade).

<Note>
  **Not built yet.** Every channel is credentialed today — the free one in step 3 is [issue #9](https://github.com/Vexa-ai/vexa-delivery/issues/9), not a running service. What else is and is not proven: [what's proven, and where](tested).
</Note>

[How it works](how-it-works) · [Install](install) · [Environments](environments/kubernetes)