How to make a 3D model in stable diffusion?

You can’t create full 3D models directly in Stable Diffusion, but you can generate 2D reference images or use depth/normal map plugins to aid 3D model creation. Stable Diffusion is a text-to-image tool—prompt for multiple views (e.g., front/side of an object) or use ControlNet to make depth maps, which you import into 3D software (like Blender) to build the mesh. For beginners, try prompting for "orthographic front view of a ceramic mug, plain background" in Stable Diffusion, then use that image as a reference in Blender to trace and extrude into a 3D shape.