Help Center/ AI/ Accessing ModelArts Inference Services via Custom Domain/ Procedure/ Getting Started
Updated on 2026-07-03 GMT+08:00
View PDF
Share

Getting Started

Publish API

Go to the API Gateway console, and click API Management > APIs > Publish to expose the APIs for comsumption. This solution creates two APIs (GET and POST) covering the common cases for AI scenarios.
Figure 1 Publish API

Table 1 APIs created by this solution

Path Prefix

Method

Usage Example

/v1

GET

List Models (v1/models): Lists all structural engines active and usable by your credentials.

/v1

POST

Create Chat Completion (v1/chat/completions): The primary gateway for text generation, complex reasoning, structured JSON extraction, and tool execution.

Access the Real-time Service

  1. Create an API key on the ModelArts console. Click Real-Time Services > API Key Authorization Management > Create API key. The API key is automatically downloaded upon creation.

    Figure 2 ModelArts console

  2. Use your subdomain and try API based on your model service, such as OpenAI API chat endpoint (/v1/chat/completions).

    Figure 3 Configure the request URL and the API key

Parent topic: Procedure