New Lovable Projects Now Use TanStack Start and Server-Side Rendering

Summary

Lovable has changed the underlying architecture used for new projects. Starting May 13, 2026, new Lovable projects are powered by TanStack Start and support server-side rendering (SSR), static-site generation (SSG), and client-side rendering (CSR) on a per-route basis.

Previously, Lovable generated React + Vite single-page applications that were client-side rendered and deployed as static files. Backend-style work generally required separately deployed edge functions. The new stack uses TanStack Start with TanStack Router, Cloudflare Workers, integrated server functions, and the existing Supabase foundation for database, authentication, storage, and realtime features.

The change is automatic for new projects. Builders do not need to opt in or learn TanStack Start before creating an app. Lovable's AI has been given curated TanStack-specific rules, developed with input from the TanStack ecosystem, to help it generate the newer patterns correctly.

SSR means page content can arrive as formed HTML rather than an almost-empty JavaScript shell. Static routes such as marketing pages and blog posts can also be prerendered at deployment. Lovable says this improves crawler visibility and social previews. Existing projects are not migrated to TanStack Start, but Lovable has added crawler prerendering to existing SPA projects. Lovable reported early increases of 2.9% in organic search traffic and 98.5% in traffic from AI tools for apps using prerendering.

Server-only secrets in the new architecture are stored separately and injected into Cloudflare Workers at request time. Lovable says rotating a secret can therefore take effect on the next request without a rebuild or redeploy.

Original Source

Read the original announcement

What This Means for Lovable Users

For people creating new Lovable projects, the architecture change happens largely behind the scenes. Publishing still works through the normal Lovable interface, but the generated application now has a more capable full-stack foundation.

The biggest visible benefit for many website builders is SEO. Server-rendered or statically generated HTML is easier for search engines, social preview services, and crawlers that do not execute JavaScript reliably to understand. This is particularly relevant for public marketing sites, directories, blogs, landing pages, and other content intended to be discovered.

Developers also gain a cleaner model for server logic. TanStack server functions can live close to the components that use them instead of requiring every backend operation to be implemented as a separately deployed edge function.

Existing projects should not assume they have been converted to TanStack Start. Lovable explicitly says existing apps continue on their previous architecture. They do, however, receive crawler prerendering for improved discoverability.

What This Means for Lovable and the Broader Community

This is a significant infrastructure change because Lovable is changing what its generated applications fundamentally are. Moving from static client-rendered SPAs toward a full-stack React framework gives the platform more flexibility around rendering, server logic, secrets, and future application architecture.

It also highlights an unusual challenge for AI coding platforms: new frameworks can appear after a model's training cutoff. Lovable says it addresses this by injecting current TanStack-specific rules into the AI's context, rather than relying solely on what the base model already knows.

For the broader AI-development ecosystem, this suggests that platform-maintained framework knowledge may become as important as the underlying model when generated applications depend on fast-moving libraries.

Ways You Could Use This

Build SEO-Focused Marketing Sites

New projects can use server-rendered and prerendered routes for content intended to rank and be shared.

Build this as an SEO-focused marketing site. Use the new project's server-rendering and prerendering capabilities appropriately for public content pages. Ensure each route has unique metadata and useful HTML content available to crawlers.

Keep Sensitive API Work Server-Side

Use server functions for operations involving private API keys or privileged server logic.

Implement this integration using server-side logic. Keep the private API credential server-only, never expose it through a VITE_ variable or client bundle, and explain the trust boundary before implementation.

Create Mixed Public and App Experiences

A project can combine static marketing routes with dynamic authenticated application routes.

Structure this project with public marketing pages optimized for prerendering and authenticated application routes for user-specific data. Keep shared components consistent while choosing the appropriate rendering approach per route.

Review Existing Apps Separately

Existing apps receive crawler prerendering but are not automatically migrated to the new framework.

Audit this existing Lovable project for SEO and crawler visibility. Treat it as an existing SPA unless the project itself confirms otherwise, and do not assume TanStack Start APIs are available.

If you need help implementing, reviewing, troubleshooting, or planning a production Lovable project related to this update, expert Lovable assistance is available through PromptSites.com.

Should Existing Lovable Users Do Anything?

New projects need no action; the new stack is automatic. Existing projects also do not need to migrate and should continue working as before.

If SEO is important, it is worth rechecking important pages in both new and existing projects. For existing projects, remember that crawler prerendering is not the same as a full migration to TanStack Start. Avoid asking Lovable to use TanStack-specific server patterns in an older project unless the project's actual architecture supports them.

Key Takeaway

Lovable's TanStack Start move changes the foundation of newly created projects without requiring builders to learn a new framework. The most immediate benefits are stronger rendering options, better crawler-readable output, integrated server logic, and clearer server-side secret handling. Existing apps remain on their prior stack, so architecture-specific advice needs to distinguish between old and new projects.