On Tuesday, Google unveiled an exciting update to its vibe-coding application, Opal, introducing a feature that allows users to effortlessly create automated workflows. This new capability leverages an innovative agent within Opal, enabling users to design mini-apps that can manage and execute tasks simply through text prompts.
The feature operates using the Gemini 3 Flash model, which intelligently selects the appropriate tools to carry out various tasks. For example, it can utilize Google Sheets to keep track of information across sessions, such as maintaining a shopping list for e-commerce applications. The agent autonomously devises and organizes the next steps for tasks, enhancing user efficiency.
According to Google, these agents are designed to be interactively intuitive. If additional information is required, they will prompt users for input or present options to guide the next actions. This advancement empowers individuals without technical expertise to construct intricate workflows within their applications.
Opal was initially launched for users in the United States in July 2025, allowing anyone to create mini web applications or modify existing ones. By October 2025, the tool expanded its reach to 15 additional countries, including Canada, India, Japan, South Korea, Vietnam, Indonesia, Brazil, and Singapore. In December, Google integrated Opal into the Gemini web app, which enables users to build custom applications through a visual editor without needing to write any code.
The trend of enabling app creation through natural language prompts is not limited to Google. Numerous startups are also developing similar tools. Notable examples include Lovable and Replit, along with emerging players like Wabi, founded by a former Replika creator, and Emergent, which has gained backing from Softbank and Lightspeed. These platforms are rapidly gaining traction in the tech landscape.