It’s been an incredible few months for Firebase. We just wrapped up Firebase + Gemini Day in São Paulo, where we got hands on with the local developer community to explore app development with Firebase and Gemini. Today, we’re excited to continue connecting with the local developer community at Cloud Summit in Brazil, where we’re sharing recent updates across the Firebase platform to help you build AI-powered apps with greater speed and efficiency, including:
- More control over the design and cost of AI features through the Firebase AI Logic SDK
- Faster deployment and new insights for Firebase App Hosting, which also works better with your tools
- Improved onboarding and direct, in-console schema editing with Firebase Data Connect
- Type-safety, production ready stability with Genkit Go 1.0
Read on for more details!
More customizations over your AI features with Firebase AI Logic
With Firebase AI Logic you can build generative AI features directly into your mobile and web apps. You have shared your feedback, thank you, and we have heard you loud and clear, so we have just rolled out some exciting new features that are sure to make developing AI-powered user experiences easier:
Thinking Budget & Thought Summary Configuration
Generative AI models that are able to “think” generate significantly better responses, but it can come at a cost. To help you manage your budget and reduce latency, you can now define a thinking configuration. You can also set up your configuration to include thought summaries in the model’s output so you can examine the model’s reasoning process.
Imagen Image Editing (Android & Flutter)
It’s now easier than ever to edit and generate images from the client side using the Firebase AI Logic SDKs and Imagen models (for Android & Flutter) Simply provide a reference image and a prompt describing the required changes and Imagen will handle the rest.
Live API
With the Firebase AI Logic SDKs and the Gemini Live API you can build conversational experiences for your users. Today, we are expanding our support to include a no-cost option through the Gemini Developer API.
Hybrid Experience (Chrome)
The Firebase team working closely with the Chrome team have made it easier to use the hybrid on-device experience by supporting it in the main Firebase JS SDK. The Firebase JS SDK will automatically check for the availability of Gemini Nano in the user’s browser and based on your configuration switch between on-device and cloud-hosted models.
AI Monitoring
We’ve also added insights into the performance of the Gemini Developer API to the AI monitoring dashboard, giving you free-tier access to observability tools to help you optimize your AI implementation.
To learn more about these updates, check out the blog post.
Faster app deployments and more from Firebase App Hosting
App Hosting gives you a seamless deploy and hosting experience for your full stack web applications. Today, these updates make that experience even better:
Faster deployments
We are rolling out a series of optimizations to the build pipeline to speed up deployment. The first of these optimizations results in smaller app binaries that take less time to upload and download so you can get your app to market faster.
Route-based monitoring
Now you can identify routes within your app and the App Hosting dashboard will show metrics, like error rates, request counts and more, specific to each route. This is useful for understanding and optimize the performance of the most important parts of your app.
Improved CI/CD setup
We’ve decoupled App Hosting’s backend from your CI/CD workflow to give you flexibility to connect your GitHub repository when you are ready. For example, you can start development with Firebase Studio and deploy to App Hosting. Later at your convenience, you can create a GitHub repo for your app and connect your existing App Hosting backend later - but still automatically get all the CI/CD benefits App Hosting provides.
Gemini CLI /deploy prompt
Using the Gemini CLI, you can now ask Gemini to deploy your full-stack app to App Hosting with the new /deploy prompt in the Firebase MCP server.
To learn more about these features, check out the blog post.
Easier onboarding, schema-editing, and greater transparency in Firebase Data Connect
To accelerate your backend development cycle, we’ve made the following updates to Data Connect, Firebase’s first relational database solution built on Cloud SQL for PostgreSQL.
Improved Firebase console and CLI onboarding
Getting started with Data Connect just got easier. You can set up your schemas, seed your database and generate your initial operations via the Firebase console or the Firebase CLI.
In-Console Schema Editing
To edit Data Connect schemas, you are no longer limited to the Firebase CLI. You can now edit Data Connect schemas directly in the Firebase console.
Access Transparency
New Access Transparency logs in the Cloud console give you greater visibility and control over your data. You can now view logs in the Cloud console for greater visibility and control over data access.
Production ready stability with Genkit
Genkit, our open-source framework for building GenAI-powered applications and features, is becoming even more powerful with its first stable release for Go and enhanced support for AI-assisted development.
Genkit Go 1.0 Stable Release
The Genkit Go SDK is now stable and production-ready, offering a unified API for multiple model providers (Google AI, Vertex AI, OpenAI, Ollama), type-safe flows, tool calling, and easy deployment as HTTP endpoints.
Enhanced AI-Assisted Development
The new genkit init:ai-tools command streamlines integration with AI coding tools, like the Gemini CLI and Firebase Studio. This provides tools with the latest knowledge of the Genkit framework, allowing them to answer questions, generate accurate Genkit code, and debug AI flows by analyzing traces.
To learn more about these features, check out the blog post.
Your feedback in action
Many of the above updates we’ve made to Firebase are a direct result of your feedback, so please keep telling us how we can make our products even better through Firebase’s UserVoice.
We’re incredibly excited about the future of AI-powered applications, and we can’t wait to see what you build next with Firebase! You can stay up to date on the latest news and best practices on using all Firebase products by subscribing to our YouTube channel, reading our blog, and following us on X and LinkedIn.