3D Model - Parameters Reference
🎯 Rodin Hyper3D
Model ID: model_rodin-hyper3d
Parameter | Type | Allowed Values | Default | Description |
---|---|---|---|---|
prompt | string | - | - | Text prompt to guide generation. Required for Text-to-3D. Optional for Image-to-3D. |
input_image_urls | file_array (image) | - | - | Images to use for Image-to-3D. Required for Image-to-3D, optional for Text-to-3D. |
condition_mode | string (enum) | fuse, concat | concat | Determines how images are combined (fuse merges features, concat uses multi-view images). |
geometry_file_format | string (enum) | glb, usdz, fbx, obj, stl | glb | Output file format. |
material | string (enum) | PBR, Shaded | PBR | Material type for the 3D model. |
quality | string (enum) | high, medium, low, extra-low | medium | Quality of the generated 3D model. |
use_hyper | boolean | true, false | false | Whether to export the model using hyper mode. |
tier | string (enum) | Sketch, Regular | Regular | Tier of generation (Sketch or Regular). |
TAPose | boolean | true, false | false | For human-like models, force a T/A pose. |
bbox_condition | number_array(3) | - | - | Defines X/Y/Z dimensions of the bounding box. |
seed | number (0–65535) | - | - | Randomization seed. Leave blank for random. |
🎯 Rodin Hyper3D HighPack
Model ID: model_rodin-hyper3d-highpack
Parameter | Type | Allowed Values | Default | Description |
---|---|---|---|---|
prompt | string | - | - | Text prompt to guide generation. Required for Text-to-3D. Optional for Image-to-3D. |
input_image_urls | file_array (image) | - | - | Images to use for Image-to-3D. Required for Image-to-3D, optional for Text-to-3D. |
condition_mode | string (enum) | fuse, concat | concat | Determines how images are combined (fuse merges features, concat uses multi-view images). |
geometry_file_format | string (enum) | glb, usdz, fbx, obj, stl | glb | Output file format. |
material | string (enum) | PBR, Shaded | PBR | Material type for the 3D model. |
quality | string (enum) | high, medium, low, extra-low | medium | Quality of the generated 3D model. |
use_hyper | boolean | true, false | false | Whether to export the model using hyper mode. |
tier | string (enum) | Sketch, Regular | Regular | Tier of generation (Sketch or Regular). |
TAPose | boolean | true, false | false | For human-like models, force a T/A pose. |
bbox_condition | number_array(3) | - | - | Defines X/Y/Z dimensions of the bounding box. |
addons | string | HighPack | HighPack | Enable HighPack for 4K textures and high-poly models (triple cost). |
seed | number (0–65535) | - | - | Randomization seed. Leave blank for random. |
🎯 Tripo V2.5 Multiview to 3D
Model ID: model_tripo-v2-5-multiview-to-3d
Parameter | Type | Allowed Values | Default | Description |
---|---|---|---|---|
front_image_url | file (image) | - | - | Front view image of the object (required). |
left_image_url | file (image) | - | - | Left view image of the object. |
back_image_url | file (image) | - | - | Back view image of the object (required). |
right_image_url | file (image) | - | - | Right view image of the object. |
orientation | string (enum) | default, align_image | default | Aligns the model orientation to the original images. |
pbr | boolean | true, false | true | Enable PBR generation, ignores texture parameters. |
face_limit | number | - | - | Limits number of faces; auto-determined if unset. |
texture | string (enum) | no, standard, HD | standard | Texture quality (no texture, standard, or HD). |
texture_alignment | string (enum) | original_image, geometry | original_image | Texture alignment priority. |
texture_seed | number | - | - | Random seed for texture generation. |
auto_size | boolean | true, false | false | Automatically scale to real-world dimensions. |
quad | boolean | true, false | false | Enable quad mesh output (forces FBX). |
seed | number | - | - | Random seed for model geometry generation. |
🎯 Tripo V2.5 Image to 3D
Model ID: model_tripo-v2-5-image-to-3d
Parameter | Type | Allowed Values | Default | Description |
---|---|---|---|---|
image_url | file (image) | - | - | Image URL used for model generation (required). |
orientation | string (enum) | default, align_image | default | Aligns the model orientation to the original image. |
pbr | boolean | true, false | true | Enable PBR generation, ignores texture parameters. |
face_limit | number | - | - | Limits number of faces; auto-determined if unset. |
texture | string (enum) | no, standard, HD | standard | Texture quality (no texture, standard, or HD). |
texture_alignment | string (enum) | original_image, geometry | original_image | Texture alignment priority. |
texture_seed | number | - | - | Random seed for texture generation. |
auto_size | boolean | true, false | false | Automatically scale to real-world dimensions. |
quad | boolean | true, false | false | Enable quad mesh output (forces FBX). |
seed | number | - | - | Random seed for model geometry generation. |
🎯 Direct3D S2
Model ID: model_direct3d-s2
Parameter | Type | Default | Description |
---|---|---|---|
image | file | `` | Front view to transform in 3D, should remove background to get better results |
maxFaceNum | number | 40000 | The max number of faces to target for the generation |
sdfResolution | number | 1024 | Controls the level of detail in the 3D model. Higher values produce more detailed shapes but take longer to process. |
paint | boolean | True | Paint the mesh with texture. |
seed | number | `` | Use a seed for reproducible results. Leave blank to use a random seed. |
🎯 Hunyuan 3D v2.1
Model ID: model_hunyuan-3d-v2-1
Parameter | Type | Default | Description |
---|---|---|---|
image | file | `` | Front view to transform in 3D, should remove background to get better results |
paint | boolean | True | Paint the mesh with texture. |
steps | number | 30 | The number of steps for the generation |
guidanceScale | number | 5 | Higher values adhere more closely to the input image, while lower values allow for more creative freedom. |
targetFaceNum | number | 40000 | The number of faces to target for the generation |
seed | number | `` | Use a seed for reproducible results. Leave blank to use a random seed. |
🎯 Hunyuan 3D v2.0
Model ID: model_hunyuan-3d
Parameter | Type | Default | Description |
---|---|---|---|
image | file | `` | Front view to transform in 3D, should remove background to get better results |
paint | boolean | True | Paint the mesh with texture. |
steps | number | 30 | The number of steps for the generation |
guidanceScale | number | 5 | Higher values adhere more closely to the input image, while lower values allow for more creative freedom. |
targetFaceNum | number | 10000 | The number of faces to target for the generation |
seed | number | `` | Use a seed for reproducible results. Leave blank to use a random seed. |
🎯 Hunyuan 3D Fast
Model ID: model_hunyuan-3d-fast
Parameter | Type | Default | Description |
---|---|---|---|
image | file | `` | Front view to transform in 3D, should remove background to get better results |
paint | boolean | True | Paint the mesh with texture. |
steps | number | 30 | The number of steps for the generation |
guidanceScale | number | 5 | Higher values adhere more closely to the input image, while lower values allow for more creative freedom. |
targetFaceNum | number | 10000 | The number of faces to target for the generation |
seed | number | `` | Use a seed for reproducible results. Leave blank to use a random seed. |
🎯 Hunyuan 3D Fast Multi View
Model ID: model_hunyuan-3d-mv-fast
Parameter | Type | Default | Description |
---|---|---|---|
image | file | `` | Front view to transform in 3D, should remove background to get better results |
backImage | file | `` | Back view to transform in 3D, should remove background to get better results |
leftImage | file | `` | Left view to transform in 3D, should remove background to get better results |
rightImage | file | `` | Right view to transform in 3D, should remove background to get better results |
paint | boolean | True | Paint the mesh with texture. |
steps | number | 30 | The number of steps for the generation |
guidanceScale | number | 5 | Higher values adhere more closely to the input image, while lower values allow for more creative freedom. |
targetFaceNum | number | 10000 | The number of faces to target for the generation |
seed | number | `` | Use a seed for reproducible results. Leave blank to use a random seed. |
🎯 Hunyuan 3D Multi View
Model ID: model_hunyuan-3d-mv
Parameter | Type | Default | Description |
---|---|---|---|
image | file | `` | Front view to transform in 3D, should remove background to get better results |
backImage | file | `` | Back view to transform in 3D, should remove background to get better results |
leftImage | file | `` | Left view to transform in 3D, should remove background to get better results |
rightImage | file | `` | Right view to transform in 3D, should remove background to get better results |
paint | boolean | True | Paint the mesh with texture. |
steps | number | 30 | The number of steps for the generation |
guidanceScale | number | 5 | Higher values adhere more closely to the input image, while lower values allow for more creative freedom. |
targetFaceNum | number | 10000 | The number of faces to target for the generation |
seed | number | `` | Use a seed for reproducible results. Leave blank to use a random seed. |
🎯 Trellis
Model ID: model_trellis
Parameter | Type | Default | Description |
---|---|---|---|
images | file_array | `` | Input image(s) to generate 3D asset from |
ssGuidanceStrength | number | 7.5 | Stage 1: Sparse Structure Generation - Guidance |
ssSamplingSteps | number | 12 | Stage 1: Sparse Structure Generation - Steps |
slatGuidanceStrength | number | 3 | Stage 2: Structured Latent Generation - Guidance |
slatSamplingSteps | number | 12 | Stage 2: Structured Latent Generation - Steps |
meshSimplify | number | 0.9 | GLB Extraction - Mesh Simplification |
textureSize | number | 2048 | GLB Extraction - Texture Size |
seed | number | `` | Use a seed for reproducible results. Leave blank to use a random seed. |
Updated 5 days ago