Google announced a new AI agent for its vibe-coding app Opal that enables users to create mini apps capable of planning and executing tasks through text prompts. The feature is powered by Gemini 3 Flash and can automatically select tools to complete tasks.
For example, the agent can use Google Sheets to maintain memory across sessions, such as tracking a shopping list for an e-commerce app. It independently plans next steps and interacts with users if additional input is required, offering prompts or choices to move workflows forward. Google says the update allows non-technical users to build complex automated workflows without writing code.
Opal first launched in the U.S. in July 2025 and later expanded to 15 additional countries. In December, Google integrated Opal into the Gemini web app, enabling visual app creation and remixing.
The move comes amid growing competition in the no-code, prompt-based app development space. Startups like Lovable, Replit, Wabi, Emergent, and Rocket.new are also building tools that let users create applications using natural language. It also follows the rollout of Gemini 3.1 Pro across consumer, developer, and enterprise products, where the company has highlighted major improvements in reasoning performance and benchmark results.
