3D Gaussian splatting lets you tour homes right in your browser
An open-source workflow built on PlayCanvas allows anyone to scan an entire house with just a smartphone and explore it directly in a web browser. Instead of traditional 3D polygons, it uses millions of tiny light clouds to render photorealistic spaces in files smaller than a TikTok video. It drops scanning costs to around $200, giving freelancers a fast new side hustle.
You don't need a clunky VR headset or a massive dedicated app to tour a home anymore. Someone just scanned an entire house with a smartphone, uploaded it, and made photorealistic exploration possible right inside a standard mobile browser.
Tiny light clouds replacing heavy 3D polygons
Built on PlayCanvas and released as open source, the workflow relies on 3D Gaussian splatting. Instead of stitching together triangular polygons like traditional video games, it uses millions of oriented light clouds to represent depth and color, with AI reconstructing reality directly from smartphone photos.
The resulting file is literally smaller than an average TikTok video, allowing it to load instantly on phones. Compared to a $15,000 commission on a $500,000 apartment, scanning costs sit around $200, letting prospective buyers tour 50 listings in a single afternoon. Freelancers are already turning this into a side hustle, charging real estate firms, Airbnb hosts, and museums anywhere from $300 to $800 per scan.
Why it matters
One person, one smartphone, and one weekend is now enough to launch an actual business. By taking expensive LiDAR hardware and dedicated apps out of the equation, it shatters the cost barrier for immersive web tours.
That said, Gaussian splats are essentially clouds of visual dots rather than solid surfaces. While they look insanely real, they lack physical collision and precise dimensional data, meaning they are built for browsing rather than replacing structural architectural tools anytime soon.
Sources (1)
- blog.playcanvas.com(opens in a new tab)Originalblog.playcanvas.com