Feature: Support GoogleGenerativeAIEmbeddings using application default credentials

[bkarduck]

Checked other resources

• This is a feature request, not a bug report or usage question.
• I added a clear and descriptive title that summarizes the feature request.
• I used the GitHub search to find a similar feature request and didn't find it.
• I checked the LangChain documentation and API reference to see if this feature already exists.
• This is not related to the langchain-community package.

Feature Description

I would like langchain to support the use of the Application Default Credentials json as the access pattern for google models when using GoogleGenerativeAI features. It currently only accepts an API key.

Use Case

I have an application through my work that uses an Application Default Credentials json file to access the google endpoint for embeddings. As vertex gets officially phased out, we're moving to using all of the google genai capabilities and while we have been using langchain's VertexAIEmbeddings wrapper to handle embedding and storing docs in elasticsearch, we want to use GoogleGenerativeAIEmbeddings instead. We found today that it's not supported unless we pass in an API key, which doesn't align with the access pattern we've been using.

Proposed Solution

I think this could be fixed by allowing the user to pass in the client they are using that has the appropriate google access with the ADC.

Alternatives Considered

The alternative would be effectively copy/pasting the class and using the same features with the correct access pattern.

Additional Context

No response

[garciasces]

I have been checking the docs, and I've found no reference to the vertex package being phased out. Do you have any link about that information?

[bkarduck]

Hi @garciasces I should clarify - it’s specifically the Generative AI module in Vertex SDK that will be deprecated after june 2026.
https://cloud.google.com/vertex-ai/generative-ai/docs/deprecations/genai-vertexai-sdk

[garciasces]

Thank you so much for your response @bkarduck , I have checked the deprecation note and passed it to my colleagues so that we can migrate our code to the Google Gen AI SDK.