Overview
Request Workflow
The Generation APIs are time and resource consuming, so we designed them to be asynchronous. This means that you submit a task without getting the result immediately.
A typical workflow involves submitting a task to the Rodin Generation endpoint, polling for its status, and downloading the generated files once the task is complete. Refer to the Minimal Example section for an example script for the workflow.
Rodin Sketch
Rodin Sketch provides fast generation of 3D models, suitable for early-stage prototyping. It produces models with basic geometry and texturing, allowing for quick visualization and iteration. The textures can be requested in Shaded or PBR formats. The model generated will only be in GLB format, thus it will have triangular meshes.
Rodin Regular
Rodin Regular generates 3D models with customizable polygon counts and high-quality texturing. It is ideal for projects that require a good balance of detail and efficiency. The textures can be requested in Shaded or PBR formats. The model generated can be in a variety of formats (ex. GLB, OBJ, FBX, etc.), and it will have triangular or quad meshes depending on the model format.
High Pack Option
For Rodin Regular, the High Pack option can be utilized for even higher resolution (4K) textures and high-poly models. High-poly model is a higher accuracy model, and the number of the high-poly faces is about 16 times the number of general faces.
Image-to-3D and Text-to-3D
For Rodin Generation, depending on the task parameters you submit to Rodin Generation endpoint, different generation modes are selected. When you upload image files, Rodin selects the Image-to-3D mode. And when you don't upload any image, you have to upload a text prompt, thus the job will be executed in Text-to-3D mode.
Multi-Images Generation
For Rodin Image-to-3D Generation, when you upload more than one image to Rodin Generation endpoint, it will automatically switch to multi-images generation mode.
You can choose the parameters you submit to Rodin Generation endpoint to change the mode of multi-image generation:
fuse
mode - if you are uploading images of multiple objects, fuse mode will combine all the features of all the objects from the images for generation.concat
mode - if you are uploading images of a single object, concat mode will inform the Rodin model to expect these images to be multi-view images of a single object.
Capabilities Comparsion
Rodin Sketch
Quick, low-resolution 3D asset generation for prototyping
~10 seconds
Basic geometry + 2K texture, simple UV mapping, low-poly
None
Rodin Regular
Detailed 3D asset generation with customizable quality
~40 seconds
Adjustable polygon counts, high-quality textures (2K texture resolution for Base Pack)
High Pack (4K textures & high-poly)
Last updated
Was this helpful?