diff --git a/apps/landing/src/components/Header.svelte b/apps/landing/src/components/Header.svelte
index c6d7004..b9f50a6 100644
--- a/apps/landing/src/components/Header.svelte
+++ b/apps/landing/src/components/Header.svelte
@@ -9,7 +9,9 @@
 
   import { SITE_URLS } from "@bashbuddy/consts";
 
-  const BANNER_STORAGE_KEY = "bashbuddy-qwen-banner-closed";
+  // const BANNER_STORAGE_KEY = "bashbuddy-qwen-banner-closed";
+  const BANNER_STORAGE_KEY =
+    "bashbuddy-banner-better-cloud-models-and-upcoming-agent-mode";
 
   let isMenuOpen = $state(false);
   let isBannerVisible = $state(false);
@@ -99,16 +101,17 @@
 {#if isBrowserEnv && isBannerVisible}
   <div
     bind:this={bannerElement}
-    class="fixed top-0 left-0 right-0 z-50 flex items-center justify-center bg-primary w-full h-min py-2 px-2 text-center"
+    class="fixed top-0 left-0 right-0 z-50 flex items-center justify-center bg-gradient-to-r from-orange-500 to-red-500 w-full h-min py-2 px-2 text-center"
     transition:slide={{ duration: 200, axis: "y" }}
   >
     <p>
       <span class="font-medium">
-        We've added Qwen 2.5 models to BashBuddy Local!
+        We're introducing better AI models on BashBuddy Cloud and upcoming agent
+        mode.
       </span>
 
       <a
-        href="/blog/qwen-2.5-models-arrive-to-bashbuddy"
+        href="/blog/better-cloud-models-and-upcoming-agent-mode"
         class="underline font-bold">Read more ➡️</a
       >
     </p>
diff --git a/packages/api/src/router/chat.ts b/packages/api/src/router/chat.ts
index eb81dcf..c4bc528 100644
--- a/packages/api/src/router/chat.ts
+++ b/packages/api/src/router/chat.ts
@@ -76,7 +76,7 @@ class GroqLLM implements LLM {
 
     const generation = trace?.generation({
       name: "chat.ask",
-      model: "llama-3.1-8b-instant",
+      model: "llama-3.3-70b-versatile",
       input: messages,
       modelParameters: {
         temperature,
@@ -87,7 +87,7 @@ class GroqLLM implements LLM {
 
     const chatCompletion = await this.groq.chat.completions.create({
       messages,
-      model: "llama-3.1-8b-instant",
+      model: "llama-3.3-70b-versatile",
       temperature,
       max_completion_tokens: maxCompletionTokens,
       top_p: topP,