Dynamic rendering serves two different versions of the same page depending on who's asking: a fully client-side-rendered version for real visitors' browsers, and a pre-rendered, fully-formed HTML version for search engine crawlers and other bots. It sits as a workaround layer in front of a JavaScript-heavy site rather than changing how the site itself works.
The approach exists because client-side rendering can be unreliable for crawling. Even though modern crawlers do execute JavaScript, that process adds cost and complexity, and dynamic rendering sidesteps it entirely for the crawler's sake by handing over content that's already fully rendered. Prerender.io and headless-browser tools like Puppeteer are common ways to generate that pre-rendered version on the fly. Google has described dynamic rendering as a legitimate interim solution rather than cloaking, as long as both versions show the same content, but it's generally treated as a stopgap: server-side rendering or static generation solves the same problem more directly, without needing to maintain two rendering paths side by side.
