Skip to main content
Preset models are public models maintained by the platform. You can filter preset models by region, cluster, and name; view their version, type, storage path, tags, and description; and use available models for subsequent development, training, or inference deployment.

​
Find Preset Models

  1. Go to Models > Preset.
  2. Select a region and cluster.
  3. To locate a specific model, enter keywords in Name.
  4. Click Search.
When locating a model, verify the model name, version, type, storage path, and target cluster.

​
View Model Details

  1. Find the target model in the preset model list.
  2. Click Details.
  3. View the model version, type, storage path, tags, description, and metadata.
When viewing details, verify that the Region and Cluster where the model resides are consistent with the subsequent workload. If you need to reference the model in Develop, Training, or Inference, also confirm the Model Name, Version, and storage path.

​
Deploy a Model

In the Preset Model list, click Deploy to proceed to the model deployment workflow.
  • The instance resources and resource pool required for the inference service have been prepared.
  • The target region and cluster can access the model.
  • The model type, model version, and runtime parameters used by the inference service have been confirmed.
To publish a preset model as an online service, see Create an LLM Inference Service.