🚀 Major upgrades just landed in SuperSplat, the free and open source platform for 3D Gaussian splatting.
Here is a 24-MILLION-Gaussian scan streaming live to a web browser. Near instant load time. Solid 60 fps.
How? A new compute-based WebGPU renderer + automatic LOD streaming. 🧵
Bit by bit, the world is being scanned — a crowd-sourced, next-gen @googlemaps you can step inside.
Real places as 3D Gaussian splats in SuperSplat by @playcanvas. Today's destination: St. Stephan's Church, Bamberg — a thousand years of history. ⛪️
WebGL Water by Evan Wallace melted brains back in 2011. 🤯
15 years on, it's back as WebGPU Water! 💧
Native WGSL shaders, real-time caustics, raytraced refraction. All in a browser tab, running on @playcanvas. 🧵
Voxel collision in SuperSplat is a game changer. 🧊
It turns any Gaussian splat into accurate 5 cm voxels, automatically. It's super-compressed and far more efficient than a mesh, yet solid enough to climb stairs and walk crossbeams.
Free & open source from @playcanvas 👇
🚀 PlayCanvas Engine 2.20 is out!
Gaussian Splatting leveled up! Relighting, soft shadows, depth of field, clipping & billions of splats, even in VR under WebGPU.
Plus procedural skies + physics joints & ragdolls.
Free and open source. ✨
npm install playcanvas
[1 / 2]
Somebody uploaded this to SuperSplat and I can barely believe it's a real place. The architecture is wow. 😲 Now anyone can visit it in their browser.
438MB of splat data, rendered in @playcanvas.
Real-time cultural heritage belongs on the web. 🧵
Oh my! Someone uploaded a HUGE 880MB splat of Baishihu Suspension Bridge to SuperSplat. It's so big - it takes 2 minutes to speedrun end to end. Largest walkable splat yet?
Rendered in-browser on the WebGPU-powered @PlayCanvas Engine. 🔥
[1 / 3]
This insanity continues. 🤯
Somebody turned a sprawling @Minecraft world into a 3D Gaussian splat.
While the graphics are now 3DGS, the collision system is still voxels! 🧊
Rendered by the @PlayCanvas Engine on WebGPU.
At 3D Discovery, we develop interactive 3DGS experiences for real estate, hospitality, and cultural heritage.
By combining 3D capture, visualization, and storytelling, we help transform places into engaging digital experiences.
Powered by @PlayCanvas.
Photoreal art galleries, streamed straight to your browser. 🖼️
No flight. No ticket. No velvet rope. Just open a link and step inside. 🔗
This is how we democratize art: by making the world’s most beautiful spaces instantly accessible to anyone, anywhere.
#playcanvas #supersplat
So hyped for this! @playcanvas now natively supports our new LCC2 format in splat-transform 2.6.0.
If you’re building large-scale, heavy 3DGS workflows, LCC2 was built to give you Dual-tracked LoD architecture and total editability. Show us what you create!
Made this a while back, clustered forward decals in @playcanvas. Allows you to project full PBR materials onto any surface at runtime. I've known for sometime playcanvas had the necessary underpinnings (these are essentially specialised form of cookie lights).
This isn’t video. It’s a photorealistic 3D Gaussian splat of a 15th–16th century church near Bergamo. 🇮🇹
Half a GB of data, but it loads instantly in the browser (thanks to the @PlayCanvas Engine).
Digital heritage and virtual tourism are about to change forever.
splat-transform 2.6.0 is here! 🚀
Free and open source library and CLI for processing 3DGS
⚡️ Faster writing of SOG format
⏩ Read @XGRIDS_OFFICIAL new LCC2 format
Release notes:
https://t.co/4tAsW4NP9s
We scanned the existing property using Gaussian Splatting, rebuilt the entire renovation project in 3D, and brought it to life in @playcanvas . Explore the project floor by floor and step inside the future home through an immersive interactive experience.
A fun way to try the new Claude Fable. I asked it to train a bot using RL to navigate 3D worlds created with @SpAItial_AI.
This is all running in the browser including RL training! Link to demo and code in comments
What you can do:
-> Train a bot from scratch, watch it learn in minutes, then drop it into a world it has never seen and it keeps navigating obstacles.
-> The worlds are real scenes generated from a text prompt or a photo via the @SpAItial_AI API (you get a splat + a collision mesh).
-> Mesh is made into a 2D walkability + height grid that the RL environment runs on.
-> The trick to generalization: the policy only ever sees egocentric inputs. A 16-ray lidar, the goal's body frame, and its speed. No global coordinates, no map.
So it learns navigation, not a memorized world.
Gaussian splats are captures frozen in time — not anymore. ⚡
Experimental RELIGHTING of splat scenes in the @playcanvas engine, driven by a proxy mesh: swap the sky, drag the sun with its soft shadows, drop in point lights — all live on a captured scene.
🔗 Runnable demo👇