Point-E is a machine-learning, text to image tool that creates 3D objects from text prompts. Describe your object and the AI will create it in one or two minutes. Objects are cloud point based and the process differs from text-conditional 3D generation, having a faster outcome.

Our method first generates a single synthetic view using a text-to-image diffusion model, and then produces a 3D point cloud using a second diffusion model which conditions on the generated image.

Learn more about the project here.

Made by Greg Brockman (Open AI).