Overview
The landscape of AI-powered 3D asset generation has evolved. The workflow is now centered around our most powerful engine yet: Rodin Gen-2.
Here is everything you need to know about the updated workflow, the robust new features, and how to utilize the legacy options effectively.
Rodin Gen-2: The New Standard
Rodin Gen-2 is now the flagship model for 3D generation on Scenario. It is designed to handle complex geometries and textures with greater accuracy, serving as the all-in-one solution for high-quality assets.
Key Features & Settings
Rodin Gen-2 introduces precise controls to tailor your output for production pipelines:
T/A Pose: A crucial feature for character artists. By toggling this on, you can force the generated character into a standard T-Pose or A-Pose, making the rigging and animation pipeline significantly easier.
Use Original Alpha: When using Image-to-3D, this toggle allows the model to respect the transparency of your input image, ensuring clean cutouts and silhouettes.
Rodin Gen-2 supports multi-image generation by default, just like the Gen-1 Highpack model. When multiple images are submitted, the model switches to multi-image generation.
Quality (Mesh Density) - Available for All Models
Regardless of whether you are using Gen-2 or Gen-1, understanding mesh density is vital for your use case.
Quality (mesh density)
This parameter defines the generation quality. Possible values are:
High: 50k faces
Medium: 18k faces (Default)
Low: 8k faces
Extra-low: 4k faces
Below are variants of the same 3D character, generated with "extra-low" to "high" quality (higher face count). These models were created using Rodin GEN-1.
Rodin Gen-1 Options
While Gen-2 is the superior choice for most high-quality tasks, the previous generation remains available as Rodin Gen-1 and Rodin Gen-1 (Highpack).
Multi-Image Support (Rodin Gen-1 Highpack)
When multiple images are submitted using Rodin Gen-1, the model switches to multi-image generation. Two configuration modes are available:
Concat Mode (Default): Use when images depict different views of the same object. This mode helps Rodin reconstruct a consistent 3D representation from various angles. One or more images are required (you can upload multi-view images in any order, regardless of the order of view).
Fuse Mode: If you are uploading images of multiple objects, fuse mode will try to combine features from the different objects in images into one single generated model.
How to Generate
The workflow remains intuitive but powerful:
Select Your Model: Choose Rodin Gen-2 for the latest technology or Rodin Gen-1 for legacy multi-image features.
Input: Describe your object in the prompt box or use the Image-to-3D feature.
Configure Settings:
Select your Quality (e.g., Medium/18k faces or High/50k faces).
If using Gen-2, enable T/A Pose for characters.
If using Gen-1 Highpack with multiple images, select Concat or Fuse mode.
Generate: Click the button to create your 3D asset.
With Rodin Gen-2 leading the way and Rodin Gen-1 providing specialized multi-image support, the 3D generation process is now faster, cleaner, and more professional.
Conclusion
The transition to Rodin Gen-2 marks a significant leap forward in AI 3D generation. By combining high-density 18K/50K quad topology with advanced features like T/A Pose enforcement and standard Multi-Image support, creators can now produce production-ready assets with greater speed and precision.
Whether you are using the cutting-edge Gen-2 for new characters or relying on Gen-1 for stylistic continuity, these tools provide the flexibility needed for modern 3D workflows.
Was this helpful?