Quickly Calling Preset Models

Constraints

The preset models and one-click model deployment functions are only available on the new console in AP-Singapore, CN-Hong Kong, and ME-Riyadh.

Introduction to Preset Models

ModelArts provides open-source models. On the ModelArts console, choose Asset Management > Models > Preset Models to view the supported models, model types, supported operations, supported resource types, and number of accelerators.

Click the target model card. On the model details page, view the model details, such as the basic information, supported capabilities, and inference features. You can select a proper model for training and inference based on the information and integrate it into your enterprise solution.

The models available in each region may vary.

Figure 1 Preset Models

Figure 2 Model details page

Supported Models

On the ModelArts console, choose Asset Management > Models > Preset Models. On the model card, click Deploy to deploy the target model. For details, see Calling a Preset Model. For details about model capabilities, see Introduction to Preset Models.

The models available in each region may vary.

Calling a Preset Model

After deploying a model on ModelArts, you can perform inference by calling the model via its API. Once you have obtained an API key, you can use cURL or code to trigger model requests. Procedure:

1. Deploy a model with one click.
   1. Log in to the ModelArts console, choose Asset Management > Models > Preset Models from the left navigation pane, and click Deploy on the model card.

      The preset model card only displays the operations supported by the model. The operations supported by different models may vary. Check the console for details.

   2. On the Create Service panel, 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.

2. Obtain an API Key
   1. On the ModelArts console, choose Model Inference > Real-Time Inference in the left navigation pane. Then, click the API Key Authorization Management tab.
   2. In the upper right corner, click Create API Key. Enter the API key name and description, select the authorization scope as required, and click OK.

      The API key is automatically downloaded upon creation. You cannot download it manually. Keep it secure as it cannot be recovered if lost.

      • If you select Specified real-time services, go to the next step to bind the API key.
      • If you select All real-time services, skip the next step.
   3. (Optional) In the API Key Authorization Management tab, click Bind in the Operation column of the API key, select the model service you want to bind to the API key, and click OK.
3. Obtain the API URL

   This guide uses a shared gateway by default. After deployment, the API URL and token information is available on the service details page.

   • API URL: On the ModelArts console, choose Model Inference > Real-Time Inference. Click the name of the deployed service. In the Service tab, obtain the Public API URL in the Network Settings area.

     

   • Model request URL: For the Qwen3-32B-64k model, the request URL is /v1/chat/completions. For more model request URLs, see the Request URL column in the preceding model list.

4. Call the Model via cURL or Python

   You can perform inference by modifying the sample code below. Simply update the full API URL, model parameter, and API key:

   • Inference API URL: Combine the public API URL and the model request URL (obtained in 3).
   • API key: Use the key obtained in 2.
   • model parameter: Use the model parameter listed in the preceding table. The following uses Qwen3-32B-64k as an example. You can modify the code as needed.

   Python

   import requests
   import json
   
   if __name__ == '__main__':
       url = "https://***/v2/infer/***/v1/chat/completions" # Real-time service URL = Public API URL + Model request URL
       api_key = "API_KEY" # Replace API_KEY with the obtained API key.
   
       # Send request.
       headers = {
           'Content-Type': 'application/json',
           'Authorization': f'Bearer {api_key}'
       }
       data = {
           "model": "qwen3_32b", # Model
           "messages": [
               {"role": "system", "content": "You are a helpful assistant."},
               {"role": "user", "content": "Hello"}
           ]
       }
       response = requests.post(url, headers=headers, data=json.dumps(data), verify=False)
   
       # Print result.
       print(response.status_code)
       print(response.text)

   Curl

   curl -X POST "https://***/v2/infer/***/v1/chat/completions" \
     -H "Content-Type: application/json" \
     -H "Authorization: Bearer $API_KEY" \
     -d '{
       "model": "qwen3_32b",
       "messages": [
         {"role": "system", "content": "You are a helpful assistant."},
         {"role": "user", "content": "Hello"}
       ]
     }'