1 The need for Apple to step in

The Apple Vision Pro is an incredible piece of hardware. Apple’s ARKit and RealityKit provide a great software foundation for developers to build on, but …

A significant challenge for individual developers and small teams creating 3D experiences is the need for assets (e.g., characters, animations for those characters, objects like tables, etc.).

Large development companies might be able to afford dedicated asset-creation teams, but small teams cannot. Even if smaller teams contract out for help, the costs would likely be too high or risky because the visionOS market is still small and not expected to grow significantly for several years. The return on this upfront investment is unclear, which also discourages larger companies from investing heavily in an unproven market.

Mathew Ball’s The State of Video Gaming in 2025 deck illustrates the risk of betting on projected sales. Each black line shows the projected growth of AR/VR headsets, while the bars show the actual sales. In March 2020, IDC projected 80 million headsets would be sold in 2024. The actual number is about 7 million - a giant miss by a factor of 10 or more!

Projected AR/VR headset sales (lines) vs. actual sales (bars) (red arrow and text are my annotations)

Indeed, Meta’s Reality Labs loses billions of dollars every quarter, and Meta’s CTO, Andrew Bosworth, recently described Horizon Worlds—Meta’s grand vision of a metaverse—as teetering on the edge of being considered a “legendary misadventure.”

Creating 3D experiences is expensive and risky.

In the past, I purchased assets from companies like Sketchfab, but this business model has proven challenging. In fact, several companies in this space have closed or been acquired.

For the next few years, Apple will need independent developers and small teams to build the incredible apps that make Apple Vision Pro an exceptional experience for its owners. But these developers need Apple’s help in creating the 3D assets required for those apps, especially assets optimized for visionOS.

2 What Apple needs to do

Apple needs to focus on three areas: assets, asset tooling, and asset generation.

2.1 Assets

The first and most important step is for Apple to create a large library of assets optimized for Apple Vision Pro / RealityKit (e.g., low-polygon counts). These include characters, character animations, furniture, audio, and textures that developers can use as-is in their applications.

Currently, Apple’s Reality Composer Pro comes with only a tiny collection of objects, sounds, and textures—far too few to build more than proof-of-concept demos. For example, there are no ready-to-use characters.

As a rough estimate, Apple should create 50–200 human (or humanoid) characters for each of several genres (fantasy, western, modern, science fiction, etc.). Every character should be rigged with Apple’s ARSkeleton3D skeleton (or whatever rigging Apple wants to use moving forward).

Apple should also provide a library of character animations and poses (walking, running, standing, falling, etc.) similar to Adobe’s Mixamo.

Mixamo animation service

Furthermore, Apple should offer face rigging, so developers can animate faces with Apple blendShapes (or whatever rigging Apple wants to use moving forward).

In addition, Apple should build numerous other 3D assets—furniture, vehicles, plants, etc.—along with relevant audio files and textures for various genres.

2.2 Asset tooling

Apple already has tools like Reality Composer Pro, Reality Converter, and its photogrammetry sample programs. The third-party photogrammetry space is fairly strong as well. However, Apple needs to improve motion-capture options for custom character animations, especially since Apple uses its own skeleton specification.

While Apple-created animations mentioned above are helpful, developers may want to record custom motion data. Apple has provided some basic libraries and a proof-of-concept motion-capture app, but it should do more to make motion capture (including face capture) easier.

WWDC 2019: Bringing People into AR

Apple also needs to extend RealityKit so characters can move fluidly in a scene. For instance, Apple should add a variant of its move(…) function that lets a character move to a new position while applying a supplied animation (i.e., running or sneaking), synchronizing the character’s footsteps with the distance traveled.

Similarly, Apple needs better support for face capture—allowing developers to record blendShapes via ARFaceAnchors along with synchronized audio, import the results into an editor, and use the results to drive a character’s face.

By providing characters whose bodies can be animated (either from Apple’s animation library or developer-recorded motion capture) and whose faces can be driven by blendShapes, Apple would equip individual developers and small development teams with powerful tools to create immersive games and storytelling experiences for the visionOS platform.

2.3 Asset generation

While providing a robust set of pre-made assets and improved tooling is crucial, Apple should also leverage its generative AI efforts to produce assets based on text prompts—akin to a Genmoji or Image Playground but for 3D. This would allow developers to create unique characters and objects without having to rely on the same 50–200 Apple-provided assets.

For example, a user could describe a character (“tall robot cowboy” or “cute humanoid alien with purple skin“) and have Apple’s tool generate a fully rigged, low-polygon model optimized for Apple Vision Pro / RealityKit.

“tall robot cowboy” and “cute humanoid alien with purple skin” generated with Apple’s Image Playground

3 Summary

The visionOS market isn’t large enough yet to justify huge investments by major studios, making small and individual developers critical for driving early adoption. However, creating spatial experiences for visionOS requires 3D assets—particularly those optimized for Apple’s platforms with special skeletons and facial rigging.

While third-party asset marketplaces exist, they haven’t always been successful, and the offerings aren’t consistently optimized for Apple’s platforms. It can also be difficult to find a cohesive set of 3D assets that fit together in one experience.

Apple must step in to fill this gap by producing assets, tooling, and APIs that empower small developers to build original, immersive experiences for Apple’s 3D platforms.