Microsoft launches Copilot 3D globally to transform 2D images into fully rendered 3D models

Microsoft has introduced a new experimental AI feature, Copilot 3D, allowing users to convert static 2D images into fully rendered 3D models without requiring specialised design skills. Available through Copilot Labs, the tool is accessible worldwide to anyone signed into a personal Microsoft account, with no subscription needed.

According to Microsoft, Copilot 3D is designed to make 3D asset creation simple and intuitive. Users can upload an image of up to 10MB, and the AI will generate a model ready for printing or rendering. These assets can be used in gaming, animation, 3D printing, augmented reality (AR), virtual reality (VR) projects, and metaverse-related content.

For inspiration, a Microsoft-curated library of unique scenes is available. However, the tool currently only supports 2D image-to-3D generation and does not yet include text-to-3D capabilities. The company recommends using the feature via Copilot on a desktop, though it can also be accessed on mobile browsers.

While Microsoft’s launch adds a user-friendly option to the space, it is not the first such offering. Stability AI introduced its Stable Video 3D (SV3D) model in 2024, while Tencent, Meta, and Apple have also released similar 3D generation tools. Apple’s Matrix3D, launched earlier this year, offers similar image-to-3D conversion.

Microsoft says Copilot 3D assets are print- and render-ready, making them suitable for immediate use in both creative and professional workflows.