Firebase AI Logic Kotlin Multiplatform (KMP) SDK

⚠️ : This is an experimental library and may be subject to breaking changes.

Build AI-powered mobile and web apps and features with the Gemini and Imagen models using Firebase AI Logic

Table of Contents

1. Supported Platforms
2. Getting Started
3. Samples

Supported Platforms

Platform	Targets Supported
Android	androidTarget
iOS	iosArm64, iosX64, iosSimulatorArm64

Getting Started

Requirements

• A Kotlin Multiplatform project with both Android and iOS targets

Project Setup

1. Add the dependency

   • Add the following to your libs.versions.toml:

   [versions]
   firebase-ai-kmp = "0.4.0"
   
   [libraries]
   firebase-ai-kmp = { module = "io.github.seanchinjunkai:firebase-ai-kmp", version.ref = "firebase-ai-kmp" }
   

   • Then, add the dependency in your commonMain source set:

   commonMain.dependencies {
       implementation(libs.firebase.ai.kmp)
   }
   

2. Firebase Console Setup

   • Go to the Firebase Console.
   • Create a new project if you haven't already.
   • In your project, navigate to the left sidebar and click on AI -> AI Logic then open Settings.
   • Enable either the Gemini Developer API, the VertexAI Gemini API, or both.

NOTE: Using the Vertex AI Gemini API requires that your project is linked to a Cloud Billing account. This means that your Firebase project is on the pay-as-you-go Blaze pricing plan.

3. Link the native iOS Bridge

   Since the SDK depends on a native Firebase iOS framework, FirebaseAIBridge, this will need to be linked to your existing iOS project.

4. Using CocoaPods

   1. Directly as a local iOS framework
      • Follow these instructions if your Kotlin Multiplatform module is integrated directly with your iOS project as a local iOS framework. That is, Xcode is calling the embedAndSignAppleFrameworkForXcode Gradle task as part of the build. At the time of writing, if you generated your Kotlin Multiplatform project using the online wizard, this is how it's set up.
      • In this scenario, you need to specify the transitive dependency on FirebaseAIBridge in your Podfile. Add the following:

        pod 'FirebaseAIBridge', :git => 'https://github.com/SeanChinJunKai/FirebaseAIBridge.git', :tag => '0.4.0'
        

5. iOS Setup

   • Create an iOS app in your Firebase project.
   • Download GoogleService-Info.plist and place it under iosApp/iosApp directory.
   • Add the FirebaseCore depedency to your Podfile:

   pod 'FirebaseCore', :git => 'https://github.com/firebase/firebase-ios-sdk.git', :tag => '11.15.0'
   

   • Add the following initialization code in iosApp/iosApp/iOSApp.swift

   import SwiftUI
   import FirebaseCore
   
   
   class AppDelegate: NSObject, UIApplicationDelegate {
     func application(_ application: UIApplication,
                      didFinishLaunchingWithOptions launchOptions: [UIApplication.LaunchOptionsKey : Any]? = nil) -> Bool {
       FirebaseApp.configure()
   
       return true
     }
   }
   
   @main
   struct YourApp: App {
     // register app delegate for Firebase setup
     @UIApplicationDelegateAdaptor(AppDelegate.self) var delegate
   
   
     var body: some Scene {
       WindowGroup {
         NavigationView {
           ContentView()
         }
       }
     }
   }
   

   • Run pod install in the iosApp directory

6. Android Setup

   • Create an Android app in the Firebase Console.
   • Download google-services.json and place it in composeApp directory.
   • Add the following to your libs.versions.toml:

   [versions]
   googleServices = "4.4.3"
   
   [plugins]
   googleServices = { id = "com.google.gms.google-services", version.ref = "googleServices" }
   

   • In the root build.gradle.kts add the plugin:

   plugins {
       alias(libs.plugins.googleServices) apply false
   }
   

   • In composeApp/build.gradle.kts, add the plugin:

   plugins {
       alias(libs.plugins.googleServices)
   }
   

7. Import FirebaseAI

   You should now be able to import FirebaseAI in commonMain

   import io.github.seanchinjunkai.firebase.ai.Firebase
   import io.github.seanchinjunkai.firebase.ai.GenerativeBackendEnum
   

8. Quick start

// Initialize the Gemini Developer API backend service
// Create a `GenerativeModel` instance with a model that supports your use case
val model = Firebase.ai(GenerativeBackend.googleAI()).generativeModel("gemini-2.0-flash")

val prompt = "Write a story about a magic backpack."
val response = model.generateContent(prompt)
print(response.text)

Try out the sample app powered by this SDK

firebase-ai-sample - Kotlin Multiplatform sample application that showcases Firebase AI integration. Runs on both Android and iOS