Google's Opal Gains New Automation Agent, Powered by Gemini

Google announced a significant upgrade to its Opal application on Tuesday, introducing a new automation agent designed to build complex workflows from simple text instructions. This feature, powered by the Gemini 3 Flash model, enables users to create functional mini-apps that plan and execute multi-step tasks autonomously.

The agent intelligently selects the necessary tools to complete a job. For example, it could generate a persistent shopping list for an e-commerce app by utilizing Google Sheets to retain information between sessions. It operates independently, devising and planning each subsequent step. Crucially, the agent is interactive; if it requires clarification or a decision, it will prompt the user with questions or choices.

Google positions this as a breakthrough for accessibility, allowing people with no coding expertise to construct sophisticated automated processes within their apps. Opal, first launched in the United States in July 2025, is a "vibe-coding" tool for building or modifying lightweight web applications. Its availability expanded to 15 additional countries last October, including Canada, India, and Brazil. By December, it was integrated directly into the Gemini web app's visual editor.

The move places Google in a growing field of startups developing natural language app builders. Competitors include established names like Lovable and Replit, alongside newer entrants such as Wabi, Emergent, and Rocket.new, all vying to democratize software creation.

Source: TechCrunch