The Khronos Group
Social Media Links
The Khronos Group is the standards body developing KHR_gaussian_splatting, the glTF 2.0 extension that defines how 3D Gaussian splats are stored inside glTF assets.
Location
N/A
Size
N/A
Total raised
N/A
Hiring?
N/A
Platform
N/A
GPU
N/A
Pricing
N/A
License
CC BY 4.0 (specification text)
Best for
Interchange of Gaussian splats between engines and viewers
Updated
July 2026
The Khronos Group develops open standards for graphics and compute, and its glTF 2.0 format is the closest thing the field has to a neutral 3D interchange container. KHR_gaussian_splatting is the extension that brings 3D Gaussian splats into that container, defining splats by position, rotation, scale, opacity and spherical harmonic coefficients, and treating the field of Gaussians as a point cloud for storage purposes. Khronos introduced the baseline in February 2026.
The extension carries a status of Release Candidate. It is written against the glTF 2.0 specification and its Known Implementations section is empty, reading only as a note to add implementations before final ratification. That is the mechanically significant fact about the extension's current state, because Known Implementations is the table Khronos uses to demonstrate that a specification has been exercised by shipping code before it is ratified.
Three implementations exist regardless of what the table says. Cesium shipped KHR_gaussian_splatting support in CesiumJS and Cesium for Unreal in March 2026, and Babylon.js and PlayCanvas have both shipped it since. The gap between shipped implementations and a specification document that records none of them is a documentation lag rather than a technical obstacle, but it does mean the extension has been sitting at Release Candidate while the ecosystem moved past it.
The contributor list reads as a map of who cares about splat interchange. Cesium supplied four contributors including Sean Lilley and Renaud Keriven, Niantic Spatial supplied two, Esri supplied two, and single contributors came from NVIDIA, Autodesk and Huawei alongside Khronos staff and independent developers including Arseny Kapoulkine and Marco Hutter. Niantic Spatial's participation matters because it also maintains SPZ, and Esri's because it ships splats in enterprise GIS.
Khronos held a public webinar on gaussian splatting in March 2026, and the specification text is published under CC BY 4.0 in the glTF repository on GitHub.
The extension is readable now in the glTF repository.
Alternatives
Recent Headlines
Esri's Arkadiusz Szadkowski on Splats at National Scale, Formats, and Finding the Early Adopters
Splatlight Gives Gaussian Splats Self-Shadowing Under Scene Lights in Blender
NouerA's ".exe" Uses 4D Gaussian Splatting — With 21 Cameras in Seoul
Captures Studio Releases Multi Splat for Aligning, Comparing and Viewing 3DGS
Esri Adds Georeferenced Gaussian Splat Layers to the ArcGIS Maps SDK for Unreal Engine at 2.4
Lightcraft Opens the Spark Story Beta, Putting Gaussian Splat Scans at the Center of Browser Previs
NVIDIA Publishes Six Agent Skills That Drive the NuRec Neural Reconstruction Stack
SHARE3DCAM Lists Gaussian Splatting as a Deliverable of Its New C10 Backpack Scanner Series
Topodrone Releases Pocket 3D, a $1,999 Handheld SLAM Scanner That Outputs Gaussian Splats
Three.js Merges a Native Gaussian Splat Renderer for WebGPU in r186