Google’s new experimental platform, Opal, allows anyone to build AI-powered mini apps without writing code. Simply by describing what they want and using visual tools, the user is able to build a mini app from scratch. The beta version launched in late July 2025 and is now available to the US public through Google Labs.
Opal offers two ways to create applications:
Build mode: Describe your desired application in natural language, and Opal’s AI will generate the workflow for you.
Example: I typed “weather-based clothing recommendation app” into the text box. Within seconds, Opal generated a complete workflow. The resulting app displays an input field on the left where users enter their city, and the app then automatically provides clothing recommendations based on current weather
conditions.
Remix mode: Start with a pre-built template application and modify it to understand how Opal works while customising the workflow to your needs.
Example: I used their Book recommendation template as a starting point for my Theatre recommendation app. I added an additional node that displays a show poster or production image above each recommendation.
Side note: The complexity of your workflow directly impacts generation time. More elaborate workflows naturally require longer processing periods. I’ve experienced some workflows that took considerable time to complete.
What to keep in mind:
Google Opal has entered the AI app development space with an intriguing proposition: what used to take days of coding can now be built in minutes through simple conversation. The platform converts natural language descriptions into visual, node-based workflows that clearly illustrate data flow, helping non-technical users grasp programming concepts without writing code.
I’m impressed by its intuitive interface, you describe what you want in plain English and watch your idea transform into a working app. For rapid prototyping, it’s genuinely impressive. The transparency is notable too: each node’s underlying prompt is visible and editable, and workflows can be modified through conversational text or direct editing.
However, this accessibility comes with trade-offs. Users cannot access underlying code, making debugging impossible when issues arise. You’re confined to Opal’s predetermined node types with no ability to implement custom logic or export code. Both creators and end-users must authenticate with Google accounts, apps can only be shared via Google-generated URLs, and deployment is restricted to Google’s infrastructure.
Despite these limitations, Opal could become a handy tool for PMs and entrepreneurs creating quick MVPs, content generation apps, educational demos, and simple productivity tools. But it’s crucial to note that Opal remains in beta, and its position in the no/low-code market will hinge on Google’s strategic priorities, their execution velocity, and their capacity to shape app development trends.