# from PyPI
pip install csm-ai
# from source
pip install git+https://github.com/CommonSenseMachines/csm-ai.git
Initialize the API client:
from csm import CSMClient
# NOTE: replace with a valid API key
csm_client = CSMClient(api_key='6bCfF4467bXXXXXX4E6B271BeC5')
Run an image-to-3d
job:
# a) using a local image path
image_path = "/path/to/image.png"
mesh_path = csm_client.image_to_3d(image_path, mesh_format='glb')
# b) using an image URL
image_url = "https://upload.wikimedia.org/wikipedia/en/a/a9/MarioNSMBUDeluxe.png"
mesh_path = csm_client.image_to_3d(image_url, mesh_format='obj')
Run a text-to-3d
job:
prompt = "3d asset of a character head, cartoon style, low poly, front view"
mesh_path, image_path = csm_client.text_to_3d(prompt, mesh_format='glb')
Mesh formats: Choose any of ['obj', 'glb', 'usdz'] for the mesh_format
argument.
Verbose mode: Run client functions with option verbose=True
(the default) to see additional status messages and logs.