A new way to quickly generate 3D models from text. Normally, it can take multiple hours to generate one 3D model from text, but this new method can generate a 3D model in only 1-2 minutes. To do this, it first generates a picture from the text then uses a different model to turn the picture into a 3D point cloud. The 3D models produced with this method are not as good as the ones made with the normal method, but it is much faster.