Migrate to use the GA version of Vertex AI in Firebase


Vertex AI in Firebase launched to General Availability (GA) on October 21, 2024.

If you used the preview version of the Vertex AI in Firebase library (FirebaseVertexAI-Preview), then you need to make the following changes to upgrade to use the GA version of the library (FirebaseVertexAI).

Here's a high-level list of the changes required to migrate:

Enable the new required API in your Firebase project (required for everyone)

You need to enable the Vertex AI in Firebase API in your Firebase project and add it to the allowlist of your Firebase API key. This is a new API that's required to use the GA version of Vertex AI in Firebase.

  1. In the Firebase console, go to the Build with Gemini page.

  2. In the Vertex AI in Firebase card, click Get started to launch a workflow that helps you to enable this API and add it to the allowlist of your Firebase API key.

  3. You can skip the rest of the workflow in the console.

Update to the GA version of the library (required for everyone)

Follow these instructions to update your app's codebase to use the GA version of the library:

  1. In Xcode, with your app project open, update your Firebase Package to version 11.4.0 or later using one of the following options:

    • Option 1: Update all packages: Navigate to File > Packages > Update to Latest Package Versions.

    • Option 2: Update Firebase individually: Navigate to the Firebase package in the section called Package Dependencies. Right-click on the Firebase package, and then select Update Package.

  2. Make sure that the Firebase package now shows version 11.4.0 or later. If it does not, verify that your specified Package Requirements allow updating to version 11.6.0 or later.

  3. Select your app's target in the Project Editor, and then navigate to the Frameworks, Libraries, and Embedded Content section.

  4. Select the + button, and then add FirebaseVertexAI from the Firebase package.

  5. Select FirebaseVertexAI-Preview, and then press the — button.

Additional changes (required depending on features you use)

Several changes were implemented for the GA versions of the SDKs. Review the following lists for any changes that you might need to make in your code to accommodate taking up the GA-version of the SDK in your app.

  • Function calling
    If you implemented this feature before GA, then you'll need to make updates to how you define your schema. We recommend reviewing the updated function calling guide to learn how to write your function declarations.

  • Generating structured output (like JSON) using responseSchema
    If you implemented this feature before GA, then you'll need to make updates to how you define your schema. We recommend reviewing the new structured output guide to learn how to write JSON schemas.

  • Timeout

    • Changed the default timeout for requests to be 180 seconds.

  • Enumerations

    • Replaced most enum types with structs with static variables. This change allows more flexibility for evolving the API in a backward-compatible way. When using switch statements, you must now include a default: case to cover unknown or unhandled values, including new values that are added to the SDK in the future.

    • Renamed the BlockThreshold enumeration to HarmBlockThreshold; this type is now a struct.

    • Removed unknown and unspecified cases from the following enumerations (now structs): HarmCategory, HarmBlockThreshold, HarmProbability, BlockReason, and FinishReason.

    • Replaced the enumeration ModelContent.Part with a protocol named Part to allow new types to be added in a backward-compatible way. This change is described in greater detail in the Content parts section.

  • Content parts

    • Removed the ThrowingPartsRepresentable protocol, and simplified the initializers for ModelContent to avoid occasional compiler errors. Images that don't encode properly will still throw errors when being used in generateContent.

    • Replaced the ModelContent.Part cases with the following struct types conforming to the Part protocol:

      • .text to TextPart
      • .data to InlineDataPart
      • .fileData to FileDataPart
      • .functionCall to FunctionCallPart
      • .functionResponse to FunctionResponsePart

  • Harm category

    • Changed the HarmCategory to no longer be nested in the SafetySetting type. If you're referring to it as SafetySetting.HarmCategory, that can be replaced with HarmCategory.

  • Safety feedback

    • Removed the SafetyFeedback type, since it wasn't used in any of the responses.

  • Citation metadata

    • Renamed the citationSources property to citations in CitationMetadata.

  • Total billable characters

    • Changed the totalBillableCharacters property in CountTokensResponse to be optional to reflect situations where no characters are sent.

  • Candidate response

    • Renamed CandidateResponse to Candidate to match other platforms.

  • Generation configuration

    • Changed the public properties of GenerationConfig to internal. They all remain configurable in the initializer.

Possible errors related to migrating

As you're migrating to use the GA version of Vertex AI in Firebase, you might encounter errors if you haven't completed all of the required changes as described in this migration guide.

403 Error: Requests to this API firebasevertexai.googleapis.com ... are blocked.

If you receive a 403 error that says Requests to this API firebasevertexai.googleapis.com ... are blocked., it usually means that the API key in your Firebase configuration file/object doesn't have a required API in its allowlist for the product that you're trying to use.

Make sure that the API key used by your app has all the required APIs included in the key's "API restrictions" allowlist. For Vertex AI in Firebase, your API key needs to have at minimum the Vertex AI in Firebase API in its allowlist. This API should have been automatically added to your API key's allowlist when you enabled the new API via the Firebase console.

You can view all your API keys in the APIs & Services > Credentials panel in the Google Cloud console.