Metaframework Landscape — frontend bundle

The 80/20 decision rule

Astro for ~80% of content-driven sites · TanStack Start for app-like SaaS · Next.js for the industry-default full-stack

One question decides almost everything: is the page mostly read, or mostly interacted with? Content is king → Astro. The app is the product → TanStack Start (type-safe) or Next.js (default). TanStack Router is the routing layer underneath — not a full metaframework by itself.

content is king → Astro app-like SaaS → TanStack Start industry default / scale → Next.js just need routing → TanStack Router

Decision helper — answer four questions

Deterministic: the same four answers always return the same pick. Try the content-heavy, minimal-JS path → it returns Astro (asserted by the gold-check).

1. Is the site primarily content (read-heavy) or an app (interact-heavy)? 2. Client-side JavaScript budget? 3. Do you need a server (DB, auth, server functions)? 4. Maturity / ecosystem priority?

Pick answers above, then hit recommend.

Where each sits on the content ↔ app spectrum

TanStack Router
(routing layer)

Astro
(content-first)

TanStack Start
(app, type-safe)

Next.js
(app, default)

← read-heavy / staticinteract-heavy / full-stack →

The four options — cards

Status verified from official sites, Jun 2026. Deep-dives live in later phases: 🔗 astro/ phase · 🔗 tanstack-start/ overview.

Comparison table — framework × dimensions

This bundle is the map — pros/cons only, no deep dive. The 80/20 rule: Astro for content, TanStack Start for app-like SaaS, Next.js for the default. TanStack Start is still Release Candidate as of Jun 2026 — not yet 1.0 GA.