WebGL and 3D on websites, explained without the hype

A pitch deck with a spinning 3D product or a scroll-triggered animation looks impressive, and increasingly it's a feature businesses ask for rather than one a studio raises on its own. The trouble is "should our site be 3D" tends to get answered on vibes, not on an honest read of what the technology actually does, what it costs, and who it's genuinely for. WebGL earns its place on a small number of sites. On the rest, it's an expensive way to slow a page down.

What WebGL actually is

WebGL is a browser standard that lets a web page draw real-time, hardware-accelerated 3D graphics — the same kind of rendering pipeline a video game uses — without installing a plugin. It isn't a design style; it's the underlying technology that makes 3D models, particle effects and shader-driven visuals possible directly in a browser tab. Most studios don't write raw WebGL day to day — libraries like Three.js, and frameworks built on top of it such as React Three Fiber, are the actual tools. WebGL is the engine running underneath them.

What it's genuinely good at

• Product configurators — letting someone rotate, zoom or recolour an actual 3D model beats a dozen static photos for things like furniture, vehicles or made-to-order goods.
• A real "wow" moment where the website itself is part of the product — a design studio, an architecture practice, a game or film — where the experience is the pitch, not a wrapper around one.
• Spatial content that's genuinely three-dimensional to begin with, like a floor plan, a map, or an architectural or scientific model.
• Portfolio and award-bait sites where differentiation is the entire point and the audience expects to be impressed, not just informed.

What it costs you, even done well

None of this is free, and not just in dollars:

• Load time — a 3D scene has to download models, textures and shader code before anything in it is usable, which works against every Core Web Vitals metric at once.
• Device range — a flagship phone handles WebGL fine; a five-year-old mid-range Android or a locked-down work laptop can choke on it, stutter, or render a blank box.
• Accessibility — a scene that only makes sense through dragging, scrolling or hovering is a real problem for keyboard and screen-reader users, and needs a genuine fallback, not an afterthought.
• Battery and data — a heavy 3D scene on a phone, on mobile data, on a train, is a bad experience for the visitor and their data plan alike.
• Maintenance — this is custom software, not a plugin. Browser and library updates can break a scene in ways an ordinary page never breaks, and usually only whoever built it can safely fix it.

The question that actually decides it

Not "would this look cool" — almost anything looks cool in a mockup. The real question is whether the 3D element does a job a static image or a well-shot video genuinely can't, for an audience actually likely to be on a device that can run it smoothly. A furniture retailer letting someone spin a sofa and preview it in three fabrics is doing a job. A generic hero animation that exists because 3D was in the brief is decoration — and decoration that costs load time on every single visit.

Where it sits, cost-wise

As general market context rather than a quote for any specific project: a genuinely bespoke, 3D or heavily animated web experience sits well above a standard custom site, for the obvious reason that it's closer to building a small piece of interactive software than assembling a page. It's usually scoped individually rather than priced off a standard range, because the variables — model complexity, how many scenes, how much custom interaction — swing enormously from one brief to the next.

The honest bottom line

WebGL is a legitimate, powerful tool for a specific set of businesses and goals, and a genuinely poor fit for most others — including plenty who'd enjoy having it. If a studio pitches 3D at you, the right follow-up isn't "can you make it more impressive" — it's "what is this specifically doing for a visitor that a well-made image or video wouldn't," plus a straight answer on how it'll actually perform on an ordinary phone, not a developer's laptop on fast office wifi. If nobody has a concrete answer to that, it's probably decoration.