diff --git a/README.md b/README.md
index 55e5063b..28440ca7 100644
--- a/README.md
+++ b/README.md
@@ -155,8 +155,9 @@ For self-hosted VM or homelab installs, `ctl.sh` wraps the common daemon lifecyc
 > 1. **Use its models as a chat provider** (supported today): add it in
 >    **Settings → Providers** as a custom OpenAI-compatible provider with
 >    `base_url = http://127.0.0.1:8642/v1` and your bearer token.
-> 2. **Delegate the whole agent loop to an external runtime**: not how Quick Start
->    works today. Tracked in [#1925](https://github.com/nesquena/hermes-webui/issues/1925).
+> 2. **Route chat through a Hermes Gateway API server** (supported today via
+>    `HERMES_WEBUI_CHAT_BACKEND=gateway`): see [`docs/advanced-chat-setup.md`](docs/advanced-chat-setup.md).
+>    Full agent-loop delegation is not yet shipped; tracked in [#1925](https://github.com/nesquena/hermes-webui/issues/1925).
 
 ### Advanced: dynamic recall prefill & Gateway-backed chat