With the release of 26.1, we have been introduced to an exciting new feature that has the potential to transform the way applications are built in Oracle APEX. My colleague, Alastair, briefly discussed this addition in his blog, ‘The Future of AI is APEX’. But in this blog, I'll take a deeper look at what APEXlang is, explore its capabilities, and highlight how it can enhance your Oracle APEX development.
APEXlang is an application format introduced in version 26.1 that represents an Oracle APEX application as a collection of structured, human-readable files that both developers and AI can understand. It enables developers to describe an application’s structure in a form that AI agents can interpret and use effectively.
The key value of APEXlang is its ability to bridge the gap between Oracle APEX and generative AI. By exposing pages, regions, processes, navigation flows and their corresponding relationships in a structured format, Large Language Models (LLMs) can understand your application both holistically and at an individual component level. This provides AI with far richer context, allowing it to provide more accurate explanations, generate features, suggest improvements and assist with troubleshooting.
APEXlang also changes the development model. Previously, most development work was performed directly within the APEX application via the builder, with exports primarily used for deployment and diff checking. However, APEXlang now opens the door to working with code editors and AI agents outside the application, creating a more flexible and modern development experience.
For development teams, APEXlang strengthens collaboration by allowing application changes to be version-controlled, reviewed via pull requests and validated using existing workflows and tools. This helps APEX applications align more closely with modern engineering practices, including CI/CD pipelines and the use of code editing tools. The result is an app representation that is easier for developers to work with and provides the structured context needed for AI to become a genuinely effective development partner.
When exporting an Oracle APEX application as you normally would, you are now given the choice between a standard SQL file output or APEXlang. When opting for the latter, the export is represented within a zip folder, logically grouping aspects of the app into different folders and files.
Once extracted, you can use an IDE such as VS Code to view and further develop the application. Along the left-hand panel, you have the two JSON files, followed by the pages (apx files), shared components, themes and the application/page group files. When clicking into them, we can see they are displayed more clearly than the original SQL output.
We can now also utilise the AI agent feature on the right-hand panel, where it's possible to request new features, troubleshoot and ask for context using natural language. This layout makes it incredibly easy to assess your application, find issues, and add new functionality. The original SQL output still works great and continues to serve its purpose, but the addition of APEXlang really feels like a huge stride forward in Oracle APEX development.
Let's now delve into the more technical aspect of this new feature.
For this example, I have created a sample HR application and added a simple classic report screen to display the contents of the employee table. However, I need to add a faceted search region to really get the most out of the data within this page.
For this, I will export my application using the APEXlang export type and interact with the AI Agents in VS Code to prompt the creation of a faceted search region.
To do this, I need to meet the following prerequisites:
Ensure VS Code and the SQL Developer extension are installed.
Establish a database connection for my APEX schema.
Once we have verified both, we can open our folder in VS Code, right-click on our application, export it from the schema, and make any changes we need. In this case, I will type the following prompt into the AI coding agent:
"Add a faceted search with checkbox group filters for department, job and work mode."
The AI agent will then review the current app and make use of APEXlang templates to create the region I have requested, adding the necessary code to my apx file.
Once this has been completed, I can review the differences against the original code and merge them. I can then click the import button, which will save these changes to my APEX schema, ready to review within the app.
As you can see, my app now has a faceted search region, courtesy of APEXlang.
This process was remarkably straightforward and did not require any manual re-import into my APEX environment. The entire workflow of generating changes, reviewing differences and importing the updated app is managed directly in VS Code.
With the infusion of AI into modern app development, it’s easier than ever for novices to understand business needs, experiment with features, and build applications. Oracle APEX is already the world’s leading low-code tool, and with this addition, it lowers the barrier to entry for new developers even more. The introduction of APEXlang and AI interactive reports in 26.1 allows for natural language to be used instead of further SQL. This means that for APEXlang, developers won’t have to add regions, pages, and features manually; instead, they can create a prompt to explain what they want added to the app. Similarly, the interactive reports allow for filters, highlighting and adjustments to the columns, etc., so they no longer require a developer, as they can be made with a simple prompt.
Just because these features are available doesn’t diminish the need for developers to still do things themselves. Like many, I enjoy the element of creating, experimenting with features, and the process of development itself. This doesn’t take away from that at all; it just adds an extra option. The introduction of APEXlang gives developers the flexibility to choose the workflow that best suits their needs. They can continue developing in Oracle APEX Builder, which has also seen major UI updates, move to a code editor such as VS Code, or leverage AI agents to assist with development, depending on the task and their preferred way of working.
The introduction of APEXlang marks an exciting new chapter for Oracle APEX, reflecting the platform's continued evolution alongside modern development practices. APEXlang lays the foundation for new ways to build applications, improve collaboration between developers and AI, and create more flexible development workflows.
As the Oracle APEX ecosystem continues to grow, it will be interesting to see how developers adopt APEXlang and the impact it has on the way applications are designed and delivered. Whether it becomes a core part of your development process or simply another tool to explore, APEXlang is a noteworthy addition that highlights Oracle's ongoing incorporation of AI into the future of APEX development.
For more information, check out our Oracle APEX Services, or contact us today, and one of our expert developers will be in touch.