After a quick play, it feels very familiar if you’re a seasoned Firebase Studio user:
Chat with AI to build app
Live preview of app + code view
Preview on diff devices
Files AI has changed can be diff’ed or restored to earlier checkpoints
Github integration (can push but can’t pull yet)
Deploy app to Cloud Run
The UI is more user-friendly for beginners. Fewer settings to manage, feels like Google are trying to reduce cognitive load for less-technical builders.
So are Firebase Studio’s days numbered? Has Google been transitioning the codebase into AI studio? And what will keep existing Firebase users from migrating to AI studio? @kirupa@tianzi@jamesor
I have noticed one major drawback with the new Google AI Studio vibe coder - it does not yet have the capability to install 3rd party libraries. Checking the available code, it appears to be because it does not expose a package.json, while Firebase Studio will add dependencies to this file via change blocks, and the dev environment will monitor changes and install via npm.
For example, Google AI Studio Vibe Coding is incapable of installing the shadcn/ui library, instead it tried to recreate the components based on its knowledge of the library. It looks like AI Studio relies only on importmap from a CDN.
Additionally, it looks like Google AI Studio uses Firebase Console for much of the same configuration used by Google Firebase Studio.
You are correct that both are Google tools designed to help developers build applications with AI. They currently offer different approaches to development.
Google AI Studio’s “Vibe Coding” is focused on a streamlined, prompt-first experience. As you noted, it manages dependencies through methods like importmap, which is great for rapidly generating applications.
Firebase Studio is a comprehensive, full-stack cloud IDE. It provides a dedicated cloud VM with full terminal access, allowing you to bring in any framework and directly manage project files like package.json.
This gives developers a choice to select the tool that best fits their project’s specific needs, whether that’s a streamlined, prompt-first workflow or a full-stack environment with deep customization.
