Updated on 2026-09-22 GMT+08:00

Model Asset Management

Introduction to Model Assets

Model asset management is the core module of ModelArts, responsible for uniformly managing both Built-in Models and My Models. This module aims to provide standardized input objects for downstream tasks including model fine-tuning and model deployment.

Built-in Models: Popular models pre-configured in ModelArts for quick and easy deployment.

My Models: Models that have been locally imported or generated after pre-training and fine-tuning through the platform.

Scenario

Model assets are available for users to deploy and fine-tune models. On the model asset page, the primary use cases include:

Deploy services with Built-in Models or My Models with few clicks for immediate inference.
Use Built-in Models or My Models for one-click fine-tuning.

Constraints

Only the new console supports this feature.
Built-in models are read-only resources and cannot be modified by users.

Viewing Preset Models

Log in to the ModelArts console. In the left navigation pane, choose Asset Management > Models.
In the Preset Models tab, view the preset models. The models are displayed in card format. Each card includes the model name, model description, and supported operations.
The models and operations available in each region may vary.

Figure 1 Preset models
Click the target model card to view details about the model, including its version, basic information, capabilities, and training features.
Figure 2 Model details

Managing My Models

The My Models page shows locally imported models and those trained or fine-tuned on ModelArts.

Importing a Model

You can import a local model to ModelArts for training and inference to create a model service that fits your needs.

Log in to the ModelArts console. In the left navigation pane, choose Asset Management > Models. Click the My Models tab.

In the upper right corner of the My Models tab, click Import Model, configure related information, and click OK.

**Table 1** Parameters for importing a model
Parameter		Description
Basic Information	Model Name	Name of the model to be imported. Must be 2–128 characters. Only letters, digits, hyphens (-), and underscores (_) are allowed. Must start with a letter, and end with a letter or digit.
	Model Type	Supported types include text generation, image understanding, image generation, video generation, and other.
	Brand	Supported platforms include major global open-source model platforms, such as DeepSeek and Qwen.
	Model Weight File Address	Manually enter the address or click to select the OBS address of the model weight files. The OBS address must start with obs:// or / and end with a slash (/). It cannot contain double slashes (//) except in the prefix. For example, obs://bucketname/path/ or /bucketname/path/. For shared buckets from other users, you must enter the path. When ModelArts connects with IAM and you use a shared bucket path, the bucket owner must grant you access and read permissions in the bucket's ACL policy. Figure 3 Bucket ACL permissions The owner of the shared bucket must set up an agency in ModelArts to grant OBS permissions to all users. To do this, select all users for authorization and choose OBS for the function permission. For details, see Configuring Agency Authorization for ModelArts with One Click.
	Model Description (optional)	Custom model description, which contains a maximum of 256 characters.
Version Information	Model Asset Version	For the initial import, the platform sets the version number to V1 by default.
Version Information	Version Description (Optional)	Description of the version, which contains a maximum of 256 characters.

In the My Models tab, you can view information about the imported model.
Figure 4 My Models

Viewing My Models

Log in to the ModelArts console. In the left navigation pane, choose Asset Management > Models. Click the My Models tab.
In the My Models tab, you can view all models in the current workspace and the models created by the current user.
In the search box, you can filter models by model name, model type, source, or brand, or enter a keyword to search for models.
Figure 5 My Models
In the My Models tab, click the target model name to view the model details.
- Model Asset Version: On the left side of the model details page, multiple version numbers may exist for the same model. During model training, if New version is selected for Publish to Assets, the system will generate a new version number based on the source model version selected for that training task. Consequently, a single model can contain multiple version iterations.
- Basic Information: Displays the basic information about the model, including the source model, training task, model type, brand, input and output types, context length, and creator.
- Operation Record: My Models can be used for both model training and real-time inference. The operation records section displays a comprehensive log of all tasks where this model was used for training or deployment. Under the Associated Task Name column, you can click a specific entry to navigate directly to that task's detail page.

Deploying My Models

ModelArts allows you to deploy a model as a real-time service. Once the deployment is successful, you can call the model via an API.

Log in to the ModelArts console. In the left navigation pane, choose Asset Management > Models. Click the My Models tab.
In the Operation column of the My Models tab, click Deploy.
If the Deploy button is grayed out, the model cannot be deployed on the My Models tab. You can go to the Model Inference > Real-Time Inference page to deploy the model as prompted. For details, see Inference Deployment.

On the Create Service page, configure related information and click OK.

**Table 2** Parameters for creating a service
Parameter	Description	Example Value
Model	Name of the model you want to deploy.	Qwen3-32B-64k
Service Name	Name used to identify and manage the real-time service. Enter a name as prompted.	service-test
Resource Pool Type	Public resource pools and dedicated resource pool are supported. Public resource pool Public resource pool for deploying the real-time service. The public resource pool supplies shared compute clusters assigned according to job parameters. Each job operates with its own isolated resources. This option offers cost-effective and flexible solutions for tasks like development and testing. Choosing the public resource pool might leave fewer resources available because of its limits. If this happens, join the queue and wait your turn. Dedicated resource pool Dedicated resource pool for deploying the real-time service. The resources provided in a dedicated resource pool are exclusive and more controllable. Use dedicated resource pools for core production services to secure exclusive resources. To select a dedicated resource pool, create one in advance.	Public resource pool
Inference Unit	Select the hardware resource configuration for the real-time service instances.	Use the recommended value.
Auto Stop	Auto-stop timer. Default: 1 hour; maximum: 24 hours. When auto stop is enabled, the system tracks how long the service runs. It will shut down the service if the runtime goes beyond the set limit. After a real-time service is deployed, you can reset the auto stop settings. To do so, choose Model Inference > Real-Time Inference in the navigation pane of the ModelArts console, locate the target service, and choose More > Configure Auto Stop in the Operation column.	Select this option. Default: 1 hour.

For details about how to call an inference service, see Inference Deployment.

Editing Model Properties

Log in to the ModelArts console. In the left navigation pane, choose Asset Management > Models. Click the My Models tab.
In the Operation column of the My Models tab, click Edit Properties, modify the model name, model type, brand, and model description as required, and click OK.
You can also click the model name and click Edit Properties on the model details page to modify the model information.

Adding a Model Version

You can add a version only when the model source is an imported model.

Log in to the ModelArts console. In the left navigation pane, choose Asset Management > Models. Click the My Models tab.
In the My Models tab, click the target model name. On the model details page, click Add Version.
On the page for adding a version, configure the model weight file address and version description, and click OK.
The model asset version automatically increases by 1.

Deleting My Models

You can delete a model that is no longer needed. If a model is being used by a training or inference task, it cannot be deleted.

Log in to the ModelArts console. In the left navigation pane, choose Asset Management > Models. Click the My Models tab.
In the My Models tab, click the model name and click Delete. In the displayed dialog box, click OK.