You're looking at a specific version of this model. Jump to the model overview.
zylim0702 /yun-3d-2.1:2e573ad9
Input schema
The fields you can use to run this model with an API. If you don’t give a value for a field its default value will be used.
| Field | Type | Default value | Description |
|---|---|---|---|
| image |
string
|
Input image for 3D generation
|
|
| mode |
None
|
textured
|
Generation mode: 'shape_only' for 3D shape only, 'textured' for shape + texture
|
| remove_background |
boolean
|
True
|
✅ CHECKBOX: Remove background from input image before 3D generation. Recommended for better 3D shape quality.
|
| quality |
None
|
High
|
3D generation quality level - affects octree resolution and detail level
|
| simplify_mesh |
boolean
|
False
|
✅ CHECKBOX: Simplify the generated mesh to reduce face count. When enabled, use 'target_face_count' slider to control final face count.
|
| target_face_count |
integer
|
500000
Min: 100 Max: 10000000000 |
🎚️ SLIDER: Target number of faces after simplification (only used when 'simplify_mesh' checkbox is enabled)
|
| merge_by_distance |
boolean
|
True
|
✅ CHECKBOX: Merge vertices by distance before texturing (recommended for texture painting). When enabled, use 'merge_distance' slider to control merge threshold.
|
| merge_distance |
number
|
0.0001
Max: 1 |
🎚️ SLIDER: Distance threshold for merging vertices (only used when 'merge_by_distance' checkbox is enabled). Lower = more aggressive merging.
|
| smooth_surface |
boolean
|
True
|
✅ CHECKBOX: Apply surface smoothing for better quality (recommended for organic models). When enabled, use 'smooth_iterations' slider to control smoothing intensity.
|
| smooth_iterations |
integer
|
2
Max: 10 |
🎚️ SLIDER: Number of smoothing iterations (only used when 'smooth_surface' checkbox is enabled). More = smoother but less detail.
|
| smooth_shading |
boolean
|
True
|
✅ CHECKBOX: Apply smooth shading to faces for better lighting and appearance
|
| enable_decimate |
boolean
|
False
|
✅ CHECKBOX: Apply Blender's decimate modifier for advanced mesh reduction using QEM (Quadric Error Metrics). When enabled, use 'decimate_ratio' and 'decimate_mode' to control reduction.
|
| decimate_ratio |
number
|
0.5
Max: 1 |
🎚️ SLIDER: Decimation ratio (only used when 'enable_decimate' checkbox is enabled). 0.0 = no faces, 1.0 = all faces. Lower = more reduction.
|
| decimate_mode |
None
|
COLLAPSE
|
🔧 MODE: Decimation algorithm (only used when 'enable_decimate' checkbox is enabled). COLLAPSE=best quality QEM edge-collapse, UNSUBDIV=good for subdivided meshes, DISSOLVE=merges flat surfaces.
|
| output_format |
None
|
glb
|
Output file format. GLB supports textures, STL is geometry-only.
|
| num_views |
integer
|
8
Min: 2 Max: 16 |
🎚️ SLIDER: Number of views for texture painting (higher = more accurate but slower). More views = better texture quality but longer generation time.
|
| paint_resolution |
integer
|
512
Min: 256 Max: 1024 |
🎚️ SLIDER: Resolution for texture painting (higher = more detailed but slower). Higher resolution = better texture detail but longer generation time.
|
| enable_vdm |
boolean
|
True
|
✅ CHECKBOX: Enable FlashVDM for massive speedup (45x faster shape generation). When enabled, use 'vdm_steps' slider to control inference steps.
|
| vdm_steps |
integer
|
15
Min: 1 Max: 100 |
🎚️ SLIDER: Number of VDM inference steps (only used when 'enable_vdm' checkbox is enabled). With FlashVDM: 5-10 steps. Without FlashVDM: 20-50 steps. Lower = faster, higher = better quality.
|
Output schema
The shape of the response you’ll get when you run this model with an API.
{'title': 'Output'}