While Microsoft was busy updating Copilot yesterday with OpenAI's new GPT-5 model, it also quietly launched Copilot 3D. It's a free-to-use feature that can transform a regular 2D image into a 3D model that can then be used in game creation, animation, 3D printing, VR / AR, and much more.
Copilot 3D just needs a good clean 2D image, and you don't even need to use any text prompts. Microsoft recommends using images with a clear background or a strong separation between the subject and the background. I've been testing Copilot 3D today and found that it works better when an image has even lighting and a sense of depth, to help the AI model bet …
