Google is bringing new vibe coding features to AI Studio, allowing both developers and non-technical users to build working applications with prompts. The update makes the platform more accessible for rapid prototyping and direct deployment to Google Cloud Run.
Where Vertex AI focuses on experts like data scientists and machine learning engineers, Google aims to make AI application development more accessible with AI Studio. Launched last year, the platform now features an extensive vibe coding workflow that allows applications to be created using natural language and simple instructions. Users with minimal technical knowledge can quickly build and test working prototypes.
New features accelerate development process
The update brings several new features to the platform. An application gallery and model selector appear in the “Build” section. For production apps, Google introduces support for “secret variables,” which allow API keys to be stored securely.
Google calls them ‘superpowers’: modular functionalities that users can add to their prompts with a single click. They are designed to accelerate AI outputs and enable the underlying Gemini model to perform deeper reasoning, media editing, and other tasks. An ‘I’m Feeling Lucky’ button offers random prompt suggestions to stimulate creativity.
From prototype to production in one click
Once the basic app is in place, users can edit specific UI elements. It is also possible to instruct Gemini to customize the interface in a targeted manner. The development process ends with a single mouse click, resulting in a live URL for testing and sharing.
The workflow seems designed for situations where speed trumps complexity. Anyone who wants to test a prototype quickly or needs a simple AI application can now get started without in-depth programming knowledge. Whether the platform is suitable for production-ready applications depends on the specific use case and required functionality.
Tip: DataGemma: Will Google’s AI make other applications obsolete?