Three frameworks. Three loud communities. We shipped real production apps on all three in 2026. Here are the numbers without the religious arguments.
- Default for B2B SaaS + content sites: Next.js 16 - largest ecosystem, best Vercel integration, RSC stability is real now.
- Default for data-heavy CRUD apps: Remix - the loader/action model is unbeatable for forms-driven apps. Less magic.
- Default for marketing sites + blazing speed: SvelteKit - smallest bundle, fastest TTI, best DX for design-led teams.
Where each one wins
- 1
Next.js 16: ecosystem + stability
Every UI library, every auth provider, every CMS has a Next.js example. RSC + server actions hit a real productivity sweet spot in 2026. App Router is no longer scary.
- 2
Remix: form-heavy enterprise apps
Loaders + actions = no useState for server data, no SWR, no React Query. Just data. For ops dashboards, admin panels, and CRM-like apps, Remix is the cleanest framework alive.
- 3
SvelteKit: when bundle size matters
Marketing sites that load on 3G. Mobile-first PWAs. Anywhere you need sub-30KB JS shipped. Svelte 5 with runes finally has the productivity of React with 1/4 the bundle.
Where teams pick wrong
Choosing SvelteKit because it's 'faster'
Performance is rarely the bottleneck. Hiring is. SvelteKit talent pool is 1/10th of React. Hire pain hits hard at 5+ engineers.
Choosing Remix because it's 'cleaner'
If your app is 80% read, 20% write, the loader/action model is overkill. Use Next.js.
Choosing Next.js because it's the default
If you're shipping a 50-page marketing site with no auth, Astro or SvelteKit will be 3x cheaper to host and 2x faster.
What we ship
B2B SaaS, internal tools, AI products
Next.js 16 + Tailwind + shadcn/ui + Clerk + Prisma + PostgreSQL. We've shipped this stack 40+ times. Boring. Predictable. Fast to build.
Heavy CRUD / ops dashboards
Remix + Tailwind + Conform + Drizzle. Forms feel native. Less code. Less to debug.
Marketing sites + landing pages
SvelteKit + Tailwind + Sanity. Lighthouse 100. Loads on dial-up.
"All three are excellent. None is dying. None is the future of everything."
Production benchmarks (mid-size SaaS app)
| Metric | Next.js 16 | Remix | SvelteKit |
|---|---|---|---|
| Lighthouse perf | 98 | 92 | 99 |
| Cold build time | 2.1s | 1.8s | 0.9s |
| Bundle size (homepage) | 180KB | 210KB | 45KB |
| Time-to-interactive (3G) | 1.4s | 1.6s | 0.8s |
| Hosting cost (Vercel/CF) | $80/mo | $40/mo | $25/mo |
| Hire-ability (talent pool) | Largest | Mid | Smallest |
| Ecosystem (UI libs, auth, CMS) | Largest | Mid | Mid |
When to pick each
Next.js 16: B2B SaaS + content sites
Largest ecosystem, best Vercel integration, RSC + Server Actions hit a real productivity sweet spot.
Remix: data-heavy CRUD apps
Loader/action model is unbeatable for forms-driven apps. Less magic. Cleaner mental model.
SvelteKit: marketing sites + speed
Smallest bundle, fastest TTI, best DX for design-led teams. Lighthouse 100 is the default.
All three: TypeScript + Tailwind
Stack underneath barely matters - the same primitives work across all three.
We've shipped on all three in production.
1-week assessment + a migration plan. Honest recommendation, not framework loyalty.
Pick by problem, not by Twitter
- Default to Next.js unless you have a specific reason to deviate.
- Choose Remix when forms + data mutations dominate your app.
- Choose SvelteKit when bundle size + load speed are the bottleneck.
- Hiring market should be a top-3 decision factor.
- All three converge on similar capabilities. Differences narrow yearly.
Pick by problem, not by Twitter
All three are excellent. None is dying. None is the future of everything. The question isn't 'which framework' - it's 'what kind of app, what hiring constraints, what perf budget'.
Default to Next.js unless you have a specific reason. The hiring market is the deciding factor for most teams.
The IRPR engineering team ships production software for 50+ countries. Idea → Roadmap → Product → Release. 200+ products live.
About IRPR