Start in the browser¶
Goal: create a model and download its generated code. You only need a browser; installation and an AI key are not required for this path.
Choose Model it to start with the visual canvas.¶
Open editor.besser-pearl.org.
Choose Model it for the visual canvas. On later visits, use File > New Project and select Low-code.
Name your project, choose Data Modeler, and click Create Project.
The project dialog lets you choose the name, view, and modeling perspective.¶
Add a class named
Bookwith an attributetitle: str.Click Quality Check and resolve any errors it reports.
Open Generate > OOP > Python Classes. Your download contains the class described by the diagram.
Use File > Export Project to save a JSON copy you can import later.
Double-click a class to edit its name and attributes in the properties panel.
This example adds a pages: int attribute as well as title: str.¶
Important
Projects are saved in this browser’s site storage. Export work you want to keep before clearing site data or switching browsers.
Want a complete walkthrough with two classes and an association? Follow Your first project in the editor guide.
Describe a system instead¶
Choose Describe it or the Agentic interface to open the assistant. Try: “Create a library model with books and authors. Each book has a title and belongs to one author.” Review the diagram before requesting code.
The assistant creates and edits models. A separate Spec-Driven Agent run generates and changes application files when you explicitly ask for it.
Next steps¶
Editor user guide: projects, diagrams, the assistant, and exporting.
Choose a generator: pick an output for your model.
Build an application with AI: generate an application from a description.