Struggling with AI app development? At the Google Cloud Next conference, Firebase announced a series of major updates—including the brand new Firebase Studio, powered by Gemini—designed to simplify AI app development and accelerate time-to-market. Let’s take a look at how Firebase is evolving to help developers tackle the challenges ahead.
The world of developers is changing fast, isn’t it? Especially with the rise of generative AI, which not only transforms how apps are built but also expands the boundaries of what’s possible. But with that innovation comes increased complexity—developers often find themselves wrestling with various emerging technologies, and the pressure adds up.
Meanwhile, businesses of all sizes are seeking more efficient AI development cycles to deliver high-quality software faster and stay ahead of the competition.
At the recently concluded Google Cloud Next conference, Firebase introduced a compelling suite of new features aimed at making it a true end-to-end platform that accelerates the entire application development lifecycle. The highlight? The debut of Firebase Studio.
Imagine a cloud-based development environment embedded with powerful Gemini AI, bringing everything you need to build and deploy production-ready AI apps into one place. That’s the core idea behind Firebase Studio, now available in preview for all developers.
Firebase Studio merges several previously released services—such as Firebase Gemini, Genkit, and Project IDX (based on Code OSS)—with existing Firebase tools and Gemini’s creativity, offering a native agentic development experience.
New to AI development? Get started with ease!
For new projects, developers can choose from over 60 prebuilt templates or begin with the App Prototyping agent. This intelligent agent lets developers express design ideas using natural language, images, drawing tools, or even screenshots—covering UI, API structure, and AI workflows.
Prototypes can be refined through iterative prompts and, once ready, deployed directly to Firebase App Hosting. Deployment generates a URL, making it easy to share a fully functional prototype for feedback or testing. A dashboard provides quick usage insights, and deeper monitoring is available via the Firebase Console.
At any point, with a single click, you can open the app in Firebase Studio’s code workspace—no setup required. Developers can tweak the architecture, expand features, and get ready for production deployment.
The code workspace offers a wide range of useful capabilities:
During the preview period, Firebase Studio offers 3 free workspaces. Google Developer Program members can access up to 30 workspaces. Try Firebase Studio now!
Firebase Studio will also offer early access to Gemini Code Assist agents. These agents help with specific tasks, such as:
AI Testing Agent available via Firebase Studio
Additionally, the unified mobile app testing service Firebase App Distribution introduces a new feature: the App Testing Agent. This agent can simulate real user interactions.
For example, if the test objective is “find a trip to Greece,” the App Testing Agent uses Gemini to plan the necessary steps, interacts with the UI on virtual or physical devices, and returns detailed pass/fail results with intuitive explanations and a visual path of the agent’s actions. Currently in preview, this feature is available to Android developers, with support for more platforms expected later this year.
Integrating cutting-edge AI features into apps often lacks standards and best practices. That’s why Firebase is continuously investing in robust frameworks, SDKs, and tools that simplify development and allow developers to focus on building engaging and innovative user experiences.
Expanded language support for Genkit
Genkit reduces the complexity of building, testing, and monitoring AI functionality in apps. It supports structured output, tool invocation, human-in-the-loop interaction, RAG, Model Content Protocol (MCP), and multi-model orchestration—ideal for developing powerful agentic experiences.
Firebase is now introducing early support for Python in Genkit and expanding support for Go, enabling developers to use their preferred languages. Access Gemini, Imagen 3, and more models like Llama and Mistral via Vertex Model Garden. You can also use self-hosted models through Ollama and tap into an expanding third-party ecosystem via community plugins.
Access new models through Vertex AI in Firebase
Vertex AI in Firebase offers a streamlined, secure SDK to integrate generative AI into apps. Thousands of apps already use this service—including meal planning app Meal Planner, AI journal assistant Life, offline hiking maps HiiKER, and creator community Waveful.
Following March’s addition of the Imagen 3 family (Imagen 3 and Imagen 3 Fast), Firebase now supports the Gemini 2.0 Multimodal Live API. This allows for natural in-app conversational experiences—for instance, enabling users to ask questions and receive answers via voice—bringing image generation and real-time multimodal capabilities to Android, iOS, Flutter, and Web apps.
Beyond AI features, Firebase is giving developers more control over app architecture and deployment through the general availability of Firebase Data Connect and Firebase App Hosting.
Build complex apps with Firebase Data Connect
Firebase Data Connect combines the stability of Google Cloud SQL for PostgreSQL with real-time GraphQL APIs and type-safe SDKs. It’s ideal for building complex apps such as:
New features in Data Connect include:
Simplify deployment with Firebase App Hosting
Firebase App Hosting is a Git-centric hosting solution for modern full-stack web apps. It manages the entire stack—from builds and CDN to server-side rendering—speeding up time-to-market. Just push your code to GitHub, and App Hosting takes care of the rest. It’s powered by enterprise-grade Google Cloud services like Cloud Build, Cloud Run, and Cloud CDN.
The GA version of App Hosting includes:
These Firebase updates may not only change how applications are built—but also redefine what can be built.
I gave Firebase Studio a quick try by launching one of the preset templates.
My first impression? “Wow, it’s really fast!” That definitely stood out.
Its target market and positioning feel somewhat similar to platforms like Lovable that offer in-browser editing and deployment—it’s designed to let developers do a lot within the browser.
I also noticed a feature to import existing projects, which looks quite handy for integrating ongoing work.
That said, I haven’t had time to dig into its privacy policy yet, so I held off uploading my own code for now—better to be cautious.
Overall, my first impression was very positive—mainly blown away by the speed. I’ll definitely be digging deeper into other features when I have more time.
Unify your chats with DMflow.chat—integrating Facebook, Instagram, Telegram, LINE, and web platforms. Our smart features include history saving, push notifications, marketing campaigns, and agent handovers for unmatched engagement and efficiency.
Devin 2.0 Has Arrived! AI Engineer Price Drops from $500 to $20? It’s Not Just Cheaper—Here Are t...
Don’t Just Know Cursor! Windsurf Uses AI Agents to “Command” Your Code—A Game-Changer for Develop...
Trae: The Next-Generation AI Code Editor, Unleashing Your Development Potential In today’s ra...
Cursor AI: The Smart Assistant for Programmers - Making Coding More Efficient and Intelligent Ex...
NeuralSVG: Turning Words into Magic—Let AI Draw Professional-Grade Vector Graphics for You! T...
Don’t Just Know Cursor! Windsurf Uses AI Agents to “Command” Your Code—A Game-Changer for Develop...
Zapier Launches MCP: A New Era of AI-Powered Automation AI Assistants Are No Longer Just Chatbot...