Register a model from various sources
Register a model from various sources using a unified API
Authorizations
Body
Details of the model to register
Indicates whether to create or update the registered model.
The name of the registered model
"churn-prediction-model"
Optional description of the registered model
"Customer churn prediction model trained on historical data"
Indicates whether this model is publicly discoverable. If true, users who are not project members will see this model in search results and can view basic model details.
Category of the model (e.g., "genai", "ml").
"genai"
Type of the model (e.g., "llm", "embedding", "other-genai", "ml").
"llm"
Project ID of the registered model. Required for Hugging Face models only. For MLflow models, this is ignored as the project is determined from the experiment run id.