Why static sites still win in 2026

Every couple of years someone declares that static sites are dead. Then Cloudflare adds another zero to its free tier and the whole conversation restarts.

This is the most boring possible take, but I’ll defend it anyway: for a blog, plain pre-rendered HTML served from a CDN is still the right default in 2026.

The pitch is unchanged

A static site is just a folder of HTML files. There’s no database, no app server, no container, no runtime. When someone visits, the CDN node nearest them serves a file. That’s it.

This unlocks four things that compounded together are hard to beat:

1. Speed. First byte latency is measured in milliseconds, anywhere on Earth.
2. Cost. Cloudflare’s free tier covers ~100,000 requests per day. Most blogs never leave the free tier.
3. Reliability. No origin to fall over. If the CDN is up, the site is up.
4. Security. Nothing to exploit. No SQL injection on a .html file.

”But what about…”

What if I want comments?

Add Giscus, Cusdis, or just link to a Mastodon thread. None of these change the fact that the post itself is static.

What if I want a contact form?

A single Cloudflare Worker (or Formspree, or Netlify Forms) handles this in about 30 lines of code. The blog stays static.

What about search?

Pagefind generates a static search index at build time. Searches happen entirely in the browser. No server.

What about real-time features?

If you legitimately need real-time features in your blog, you don’t need a blog. You need an app. Different problem.

Where I’d reach for something heavier

I’d absolutely reach for SSR (or even SSG with islands) when:

• Pages need to be personalized per user (dashboards, account pages).
• Content changes faster than I can rebuild (live scoreboards, ticker data).
• I have meaningful client-side interactivity that needs hydration.

For everything else — and especially for a blog — start static. You can always add a Worker later if you actually need one.