Google drops Opal: Make your own AI mini-apps on Gemini web
Google just rolled out Opal on the Gemini web platform, allowing users to build and share simple AI mini-apps, which are surfaced as Gems within the Gemini interface, using plain English and a drag-and-drop editor.
The new Gems menu now has handy Labs presets, a Remix feature, and an Advanced Editor that turns your ideas into multi-step workflows you can preview with text or voice, as noted in additional sources.
How does Opal work?
Opal builds step-by-step pipelines by linking prompts, models, and tools—no coding needed.
You can tweak or rearrange steps visually, test how things run, and share your creations with friends through a link.
Right now, Opal is free in public beta for users with a Google account, rolling out first in the US and to select folks worldwide.
Why should you care?
Opal helps you automate stuff like data analysis or email drafting by connecting different Gemini AI models in one workflow.
It's Google's answer to OpenAI's custom GPTs and Microsoft Copilot Studio but focuses more on making single-prompt tasks easy to remix right inside Gemini.