AI application generation

Generate runnable web application prototypes from natural language descriptions or design mockups. Skip project setup and boilerplate code to accelerate development from idea to product.

Generate a web prototype

Enter your requirements in the chat box to generate project code and preview it in real time.

Go to the Operation Orchestration Service console and click Start AI Coding.

A RAM user must have the AliyunOOSFullAccess permission.
In the input box, describe the application you want to create.

You can click an example in the Get Started section to quickly try out the feature.
- Text description: Describe your application in natural language. Example: "Create a to-do list app that supports adding, deleting, and marking tasks as complete."
- Multimodal input: Upload a screenshot or design mockup. The AI analyzes the image and generates matching code.
  
  You can also upload error message screenshots. The AI analyzes the error and suggests a fix.
After you submit your requirements, the page switches to the workspace view and generates the project code.
- Chat area: Shows the conversation history and AI reasoning process.
- Code/Preview area: Click the Code or Preview tab to switch between views.
  - Code: View the complete project code, including routes, components, styles, and configuration files.
  - Preview: View and interact with the running application.

Modify with chat commands

Refine features or adjust styles by entering modification requests in the chat input box. The code updates and the preview refreshes automatically.

Examples of modifications:

UI style: "Add a toggle for dark and light modes. Use a sun icon for light mode and a moon icon for dark mode."
Business logic: "Add a dropdown menu to the task list to filter tasks by completion status."

Modify code in the editor

Modify project files directly in the Code Editor.
The preview immediately reflects your modifications.

Roll back to a previous version

Roll back to a previous version if your changes are unsatisfactory.

Above the chat input box, click the icon.
Find the desired version in the history and click Roll back to this version.

Billing

The platform provides 100 free chat sessions daily.

The free quota resets daily. View remaining sessions and usage on the Settings page.
After the free quota is exhausted, configure a Model Studio API key on the Settings page to continue. Subsequent conversations incur billing through Model Studio.