UniForge3D
Text to 3D · Text prompt or one object image
About UniForge3D
This UniForge3D Space combines several stages: an image from your prompt, generated right/back/left camera views, and a textured 3D reconstruction. Image-to-3D mode replaces the first stage with your upload and background removal.
The four views are generated or prepared before UniForge3D combines them into a GLB. The supporting image and view-generation steps call other Spaces using the visitor’s ZeroGPU allowance. The workflow therefore depends on more than the availability of this one page.
How to use UniForge3D
- Choose text-to-3D or image-to-3D and supply a prompt or reference.
- Follow any provider sign-in requirement, then run the image and camera-view stages.
- Inspect the completed 3D result and download its textured GLB.
What to know before you start
- The documented pipeline stops if it cannot obtain all four required views, including when GPU quota is exhausted.
- Generated camera angles can disagree on details. Inspect the back and sides of the completed model.
The provider controls availability, sign-in, processing, and downloads. Save any result before reloading or closing the demo.
Does each generated view use GPU quota?
The Space documentation says the supporting image and view-generation calls use the visitor’s ZeroGPU token and allowance.
