What you can publish

• HTML. A single self-contained page.
• Markdown. Turned into a clean, styled page automatically.
• PDF. Shown in a built-in viewer, right on the page.
• Jupyter notebooks. A .ipynb file renders as a live notebook.
• Full apps. A zip of a multi-file site or a built app (for example a React or other single-page app). Thryvate serves the whole thing and picks the starting page automatically.

You can keep a site as a draft until you are ready, so nothing is reachable before you publish it.

Built a real app, not just a single page? It works. Deep links like /about or /pricing keep working even when someone refreshes the page, so your app behaves the way you expect. Larger apps and asset-heavy sites are supported too.

If you build your app from code, there are a couple of small setup tips (a relative asset path and reading the base Thryvate injects). Those live in the developer docs.

For safety, a published site cannot quietly pull in code or data from other places. If your page genuinely needs to load a font, call an API, or embed content from another site, you can allow those specific sources in the site's settings (a Premium capability). Everything else stays blocked by default.

Every site gets a share link right away. To serve it on your own domain instead, see Custom domains.