Google AI Studio integrates ’vibe coding’ to create apps without writing code

Google has introduced an update to its AI Studio platform that brings the concept of 'vibe coding', a system that generates functional applications from instructions described in text. The tool uses Gemini models to interpret user requests and automatically configure the necessary connections between different services and APIs.

The system enables the creation of multimodal applications that combine various capabilities. A user can request, for example, an application that generates videos with Veo, edits images, or incorporates Google searches, and the platform handles connecting these elements without requiring manual programming of integrations.

The update includes a visual annotation mode that facilitates subsequent modifications. Instead of describing changes through text or editing code, users can point to specific interface elements and request direct adjustments, such as changing button colors or adding animations.

Google has also revamped the AI Studio application gallery, which now functions as a reference library with visual project examples. Users can explore these applications, access starter code, and adapt it to their needs. During the app generation process, the platform displays related project suggestions generated by Gemini.

The tool initially operates under a free quota system. When this limit is exhausted, it allows users to input their own API keys to continue working without interruptions, and automatically switches back to the free tier when it renews. This functionality is currently available in Google AI Studio for developers and users without programming experience who want to create artificial intelligence application prototypes.

Key points

• Google AI Studio incorporates 'vibe coding', which generates functional applications through natural language descriptions
• The system uses Gemini models to automatically connect APIs and services without manual programming
• Enables creating multimodal apps that integrate video generation, image editing, and Google searches
• Includes a visual annotation mode to modify interface elements by pointing directly at them
• The revamped application gallery offers visual examples with starter code to adapt
• Operates with a free quota and allows adding personal API keys when the limit is reached
• Designed for both developers and users without programming experience