<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Сергей Савкин</title>
    <description>The latest articles on DEV Community by Сергей Савкин (@__0e0d9aa3ee6).</description>
    <link>https://dev.to/__0e0d9aa3ee6</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3764458%2Fb6288523-7f85-4679-8c6c-924d1e6d98c1.jpg</url>
      <title>DEV Community: Сергей Савкин</title>
      <link>https://dev.to/__0e0d9aa3ee6</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/__0e0d9aa3ee6"/>
    <language>en</language>
    <item>
      <title>I built an AI Landing Page Roaster with Next.js + GPT-4o Vision</title>
      <dc:creator>Сергей Савкин</dc:creator>
      <pubDate>Tue, 10 Feb 2026 18:58:35 +0000</pubDate>
      <link>https://dev.to/__0e0d9aa3ee6/i-built-an-ai-landing-page-roaster-with-nextjs-gpt-4o-vision-25oc</link>
      <guid>https://dev.to/__0e0d9aa3ee6/i-built-an-ai-landing-page-roaster-with-nextjs-gpt-4o-vision-25oc</guid>
      <description>&lt;p&gt;Show off: Built a landing page analyzer with GPT-4o Vision + Next.js in a weekend&lt;/p&gt;

&lt;p&gt;Tech stack:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Next.js 14 (App Router)&lt;/li&gt;
&lt;li&gt;GPT-4o Vision for screenshot analysis&lt;/li&gt;
&lt;li&gt;GPT-4o-mini for free tier&lt;/li&gt;
&lt;li&gt;Tailwind CSS + Framer Motion&lt;/li&gt;
&lt;li&gt;Vercel deployment&lt;/li&gt;
&lt;li&gt;Crypto payments (USDT/Polygon)&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The AI takes a screenshot of any URL, extracts the HTML text, and sends both to GPT-4o for a full conversion analysis.&lt;/p&gt;

&lt;p&gt;Interesting technical challenges:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Had to convert screenshots to base64 (OpenAI can't fetch external URLs reliably)&lt;/li&gt;
&lt;li&gt;Two-tier model approach saves ~90% on API costs for free users&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Live: &lt;a href="https://roastpage-ai.com" rel="noopener noreferrer"&gt;https://roastpage-ai.com&lt;/a&gt;&lt;br&gt;
Source: &lt;a href="https://github.com/sergiologino/roastpage-ai" rel="noopener noreferrer"&gt;https://github.com/sergiologino/roastpage-ai&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Would love technical feedback!&lt;/p&gt;

</description>
      <category>webdev</category>
      <category>performance</category>
      <category>frontend</category>
      <category>ai</category>
    </item>
  </channel>
</rss>
