feat(vertexai): Migrate to AI and add GoogleAI support
#8931
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
🦋 Changeset detectedLatest commit: 07f68ad The changes in this PR will be included in the next version bump. This PR includes changesets to release 2 packages
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
Size Report 1Affected Products
Test Logs |
Size Analysis Report 1This report is too large (426,227 characters) to be displayed here in a GitHub comment. Please use the below link to see the full report on Google Cloud Storage.Test Logs |
hsubox76
reviewed
Apr 16, 2025
hsubox76
reviewed
Apr 16, 2025
dlarocque
changed the title
GenAI and add GoogleAI supportAI and add GoogleAI support
dlarocque
changed the title
AI and add GoogleAI supportAI and add GoogleAI support
hsubox76
reviewed
Apr 28, 2025
* [vertexai] Use json of unary-success-citations The test files have been updated to include `json` extension versions of the tests files. The old, `txt` extension files are deprecated and will be deleted eventually. * Update also the version used in the update script
rachelsaunders
requested changes
May 2, 2025
dlarocque
commented
May 2, 2025
Comment on lines
-601
to
-606
| - title: VertexAI | ||
| path: /docs/reference/js/vertexai.vertexai.md | ||
| - title: VertexAIError | ||
| path: /docs/reference/js/vertexai.vertexaierror.md | ||
| - title: VertexAIModel | ||
| path: /docs/reference/js/vertexai.vertexaimodel.md |
There was a problem hiding this comment.
These types still exist (can't remove them until a breaking change), but our doc generation tool doesn't pick them up since they're just aliases to the new types (e.g. VertexAIError is just an alias to AIError.
dlarocque
commented
May 2, 2025
Vertex AI Mock Responses Check
|
hsubox76
approved these changes
May 6, 2025
dlarocque
commented
May 8, 2025
| '@firebase/vertexai': minor | ||
| --- | ||
|
|
||
| Add support for the Gemini Developer API, enabling usage in a free tier, and add new `AI` API to accomodate new product naming. |
There was a problem hiding this comment.
Note that this isn't the final changelog that will be in the release notes. This will be changed to the final changelog entry we decide on when we write the release notes.
rachelsaunders
requested changes
May 12, 2025
rachelsaunders
approved these changes
May 12, 2025
Merged
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
API: http://goto.google.com/googleai-in-vinf-api (internal)
This PR integrates support for the Google AI API alongside the existing Vertex AI backend. To accommodate this, the SDK has been refactored to use the more generic
AInaming convention instead ofVertexAI.Changes
Google AI Support
googleAIBackend()function to configure the SDK to use the Google AI API. This enables usage with models available in a free tier.googleAIMappers.ts) to translate requests and responses between the SDK's unified format (prioritizing Vertex AI) and the specific format required by the Google AI API. This handles differences in features like safety settings (methodunsupported), citation metadata (citationSourcesvscitations), and response properties (defaulting missing safety scores).AIErrorwith codeUNSUPPORTEDor logs warnings for features/properties not available or different in the Google AI backend (e.g.,SafetySetting.method, roundingtopK,Part.videoMetadata).Firebase AI Migration
VertexAItoAI(e.g.,AI,AIError,AIModel,AIErrorCode,AIOptions).export type VertexAI = AI;,export const VertexAIError = AIError;).Backend(GoogleAIBackend | VertexAIBackend) andBackendTypetypes to abstract the backend configuration.Initialization
getAI(app?, options?)function for initializing the SDK. It acceptsAIOptionsto specify the desiredbackend.getVertexAI(app?, options?)function for backward compatibility, initializing a Vertex AI backend instance.Internal
helpers.ts) for instance identifiers based on backend type and location, allowing the component system to manage different service instances.request.ts,stream-reader.ts) to handle backend differences and use the appropriate mappers.Packaging
./aiexport path infirebase/package.jsonto allow the package to be imported withfirebase/ai.Testing
googleAIBackend,vertexAIBackend), Google AI request/response mappers, and instance identifier helpers.generateContent,countTokens) using the Google AI backend path to verify request mapping and error handling.Sample usage