The <model> specification repeatedly emphasizes format-neutral authoring through multiple model <source> elements or server content negotiation. This is core to the entire proposal. It is supposed to be a web standard, not an Apple platform feature.
Yet Apple's official WWDC 2026 examples and documentation show this:
This is not a neutral implementation. The fallbacks they show are not model formats. This is teaching developers to write Safari-only, Apple-only website code while claiming to participate in a well defined open-standards process. This is why many developers are cynical about big vendors contributing to new web features when the results can end up completely self-serving.
The part that depresses me personally is, <model> is being used to sideline WebXR, which is a core part of the immersive web we all love. The implementation of immersive-AR sessions is being ignored on visionOS and iOS even though native apps are given the capability. <model> is being positioned as the safe declarative path when the grim reality in front of us is they are using it to push their platform as the only viable option to create AR web programs. As things stand today they are essentially forcing people to author USDZ content to use WebAR on Apple-device browsers. An open standard like WebXR being shunned as an alternative is a serious concern if you could all please please recognize how incredibly lucrative this solution that forces their format would be for them. I of course love declarative web content and <model> being an option alongside WebXR immersive-ar, but we really need to be honest about the massive potential for conflict of interest here.
But USDZ is the only current implementation format
I dearly hope this is not the justification for creating these examples of pages that only support their platform. This is the entire purpose of the working group and why we are all talking openly here. As the first implementation they are supposed to be good stewards of explaining the standard in a forward-thinking way with respect for the spec and fellow browsers. This is the first impression and a critical moment for the standard. It is absolutely not an excuse to show just a simplified slice that only supports their format because it is easy. It aggressively hinders the entire interoperability process because now code breaks with future implementations and doing it the correct way is now 'more complicated', so there is future friction ahead. An implementer with bad intentions would actually see these negatives as positives because the future confusion / broken pages further solidify their control of the spec and starts things off with negative momentum around proper implementations that don't purely favor their format.
The <model> specification repeatedly emphasizes format-neutral authoring through multiple model <source> elements or server content negotiation. This is core to the entire proposal. It is supposed to be a web standard, not an Apple platform feature.
Yet Apple's official WWDC 2026 examples and documentation show this:
This is not a neutral implementation. The fallbacks they show are not model formats. This is teaching developers to write Safari-only, Apple-only website code while claiming to participate in a well defined open-standards process. This is why many developers are cynical about big vendors contributing to new web features when the results can end up completely self-serving.
The part that depresses me personally is, <model> is being used to sideline WebXR, which is a core part of the immersive web we all love. The implementation of immersive-AR sessions is being ignored on visionOS and iOS even though native apps are given the capability. <model> is being positioned as the safe declarative path when the grim reality in front of us is they are using it to push their platform as the only viable option to create AR web programs. As things stand today they are essentially forcing people to author USDZ content to use WebAR on Apple-device browsers. An open standard like WebXR being shunned as an alternative is a serious concern if you could all please please recognize how incredibly lucrative this solution that forces their format would be for them. I of course love declarative web content and <model> being an option alongside WebXR immersive-ar, but we really need to be honest about the massive potential for conflict of interest here.
But USDZ is the only current implementation format
I dearly hope this is not the justification for creating these examples of pages that only support their platform. This is the entire purpose of the working group and why we are all talking openly here. As the first implementation they are supposed to be good stewards of explaining the standard in a forward-thinking way with respect for the spec and fellow browsers. This is the first impression and a critical moment for the standard. It is absolutely not an excuse to show just a simplified slice that only supports their format because it is easy. It aggressively hinders the entire interoperability process because now code breaks with future implementations and doing it the correct way is now 'more complicated', so there is future friction ahead. An implementer with bad intentions would actually see these negatives as positives because the future confusion / broken pages further solidify their control of the spec and starts things off with negative momentum around proper implementations that don't purely favor their format.