Key Concepts

Every site Zitegen generates is pre-built, static HTML — nothing renders it per request, and no client-side framework hydrates it in the browser. That's what makes it fast to load and trivial for search crawlers to read: the full page is already there in the response, no JavaScript execution required.

No React, Vue, or any client-side framework ships to your visitors. Every element carries a data-zitegen-id back to the component tree, and specific behaviors — a modal, a filterable collection, an inline-edited section — are driven by small, purpose-built data attributes instead of a general-purpose runtime. Pages stay light because there's nothing heavier to load.

Publish deploys through that same Cloudflare adapter to Cloudflare's global edge network, live at yourslug.zitegen.app in moments. Static files serve straight from the edge — no origin server to wait on.

The component tree — a validated, structured document, not free-form markup — is the single source of truth, stored in Zitegen's own database rather than bundled into your site. AI chat, the visual canvas, and the code editor all write through that same tree, so nothing falls out of sync. Things like form submissions and articles are served through Zitegen's own API at request time, not a database shipped inside the static output.

Every site gets a generated sitemap.xml, robots.txt, and an rss.xml feed of your published articles, plus per-page title, description, canonical URL, and schema markup you can set from Site Settings. By default an llms.txt also indexes your pages for AI crawlers — the AI crawler policy in Site Settings lets you keep that, ask crawlers to skip AI training while still allowing it, or block named AI crawlers outright (which stops generating llms.txt). Because pages are static HTML, search crawlers see the finished page immediately — nothing to execute first.