HTMX and the Hypermedia Revival: Building Web Apps Without JavaScript Frameworks
March 18, 2026
In a world where every new web project defaults to React, HTMX is a quiet rebellion. It asks a simple question: what if the browser's native capabilities were enough, and we just needed better HTML?
HTMX extends HTML with attributes like hx-get, hx-post, hx-swap, and hx-trigger. These let any HTML element make HTTP requests and swap parts of the DOM with the response — without writing a single line of JavaScript. The server returns HTML fragments, not JSON. The browser handles the rest.
This sounds like a step backward. It is not. For a large class of applications — CRUD apps, admin panels, dashboards, content sites — the React model introduces enormous complexity for minimal benefit. You need a build system, a bundler, a state management solution, an API layer, and a hydration strategy. HTMX needs none of these. You write server-side templates (in any language: Python, Go, Ruby, Elixir) and sprinkle HTMX attributes where you need interactivity.
The performance story is compelling. There is no JavaScript bundle to parse and execute on page load. The initial render is server-side HTML — fast, SEO-friendly, and accessible by default. Subsequent interactions fetch only the HTML fragment that needs to change, not the entire page.
The creator of HTMX, Carson Gross, articulates the philosophy as "Hypermedia as the Engine of Application State" (HATEOAS) — a REST principle that most REST APIs ironically violate. The server drives the application state through hypermedia controls embedded in the response.
HTMX is not for every application. Real-time collaborative features, complex client-side state, and offline-first requirements genuinely need JavaScript. But for the vast majority of business applications, HTMX paired with a fast server-side framework (FastAPI, Phoenix, Rails, Hono) delivers a simpler, faster, more maintainable stack than the SPA default.
The 30,000+ GitHub stars and growing adoption in the Django and Rails communities suggest this is not a niche experiment. It is a legitimate architectural choice that deserves serious consideration.