· product · comparison

Honest comparison.

We sell Vesk, but we sell it honestly. Below is where compiler-first wins, where it ties, and where you should choose something else. Measure, don't believe.

Try Vesk → See features

head-to-head — where framework choice matters most

Pillar	Vesk	Next.js 15	SvelteKit	Astro + islands	Qwik / Remix
Paradigm	Compiler-first, VDOM-free	React VDOM + RSC	Compiler-aided, fine-grained	Static-first, islands opt-in	Resumable (Qwik) / SSR (Remix)
JS shipped	Static 0 B; app ~22-48 kB	~90-140 kB baseline	~28-45 kB	0 B → islands add React/Svelte	Qwik 0-20 kB; Remix ~40 kB
Hydration	Claim walk, static skipped; 4 strategies	Full hydrateRoot	Full + runtime	Islands only	Qwik resumable; Remix hydrate
Routing	File + manual, groups, hash, offline	File, no manual, no hash	File, good	File, simple	File
Data	useFetch Resource, SSR handoff, staleTime	fetch + RSC, implicit	load functions	fetch at build	loaders
Native	Kotlin/Compose from same IR	None (Expo separate)	None	None	None
Build error	Invalid semantics fail build	Runtime, sometimes silent	Some compile errors	Few	Mixed
Ecosystem	npm, Tailwind, lucide — small, growing	Massive	Solid	Huge (framework-agnostic)	Small (Qwik) / Remix: medium

We benchmarked on showcase apps (same UI/data) on CI 4 vCPU Node 20 gzip. Re-run on your app — architecture matters more than micro-benchmarks.

When to choose Vesk — we win

Content + forms + lists: docs, dashboards, admin, field tools — where per-cell updates and zero-JS chrome matter.
Web + native with shared design: one token set → Tailwind + Compose, not two apps.
You want guarantees: compile-time proof beats runtime hope.

When not to — be honest

Huge React ecosystem need: Next's npm reach is unmatched — if you need 10 niche React libs day one, choose Next.
Heavy canvas/WebGL: Vesk is DOM-first; for games, Flutter/Canvas is better.
Team is React-only: Vesk asks you to learn component + &[track] — small but real.

The 90-second sell — tailors to your current stack

You use Next.js

Keep file routing, lose VDOM tax. Move state to track , keep data in useFetch — most pages become static and ship 0 B. Port in a day.

Migration →

You use SvelteKit

Stay fine-grained, gain Kotlin target + manual routing + offline boundaries. Same feel, more reach.

Native →

You use Astro/Qwik

Keep islands + 0 JS, gain a real framework — routing, data, forms, auth middleware — without gluing.

Features →

Get started → See proof