`GoogleGenerativeAIEmbeddings`¶

Reference docs

This page contains reference documentation for GoogleGenerativeAIEmbeddings. See the docs for conceptual guides, tutorials, and examples on using GoogleGenerativeAIEmbeddings.

GoogleGenerativeAIEmbeddings ¶

Bases: BaseModel, Embeddings

Google Generative AI Embeddings.

Setup

Vertex AI Platform Support

Added in langchain-google-genai 4.0.0.

GoogleGenerativeAIEmbeddings now supports both the Gemini Developer API and Vertex AI Platform as backend options.

For Gemini Developer API (simplest):

Set the GOOGLE_API_KEY environment variable (recommended), or
Pass your API key using the google_api_key kwarg

For Vertex AI:

Set vertexai=True and provide project (and optionally location).

Example:

from langchain_google_genai import GoogleGenerativeAIEmbeddings

# Gemini Developer API
embeddings = GoogleGenerativeAIEmbeddings(model="gemini-embedding-001")
embeddings.embed_query("What's our Q1 revenue?")

# Vertex AI
embeddings = GoogleGenerativeAIEmbeddings(
    model="gemini-embedding-001",
    project="my-project",
    vertexai=True,
)

**Automatic backend detection** (when `vertexai=None` / unspecified):

1. If `GOOGLE_GENAI_USE_VERTEXAI` env var is set, uses that value
2. If `credentials` parameter is provided, uses Vertex AI
3. If `project` parameter is provided, uses Vertex AI
4. Otherwise, uses Gemini Developer API

Environment variables

Variable	Purpose	Backend
`GOOGLE_API_KEY`	API key (primary)	Both
`GEMINI_API_KEY`	API key (fallback)	Both
`GOOGLE_GENAI_USE_VERTEXAI`	Force Vertex AI (`true`/`false`)	Vertex AI
`GOOGLE_CLOUD_PROJECT`	GCP project ID	Vertex AI
`GOOGLE_CLOUD_LOCATION`	GCP region (default: `us-central1`)	Vertex AI
`HTTPS_PROXY`	HTTP/HTTPS proxy URL	Both
`SSL_CERT_FILE`	Custom SSL certificate file	Both

GOOGLE_API_KEY is checked first for backwards compatibility. (GEMINI_API_KEY was introduced later to better reflect the API's branding.)

Proxy configuration

Set these before initializing:

export HTTPS_PROXY='http://username:password@proxy_uri:port'
export SSL_CERT_FILE='path/to/cert.pem'  # Optional: custom SSL certificate

For SOCKS5 proxies or advanced proxy configuration, use the client_args parameter:

embeddings = GoogleGenerativeAIEmbeddings(
    model="gemini-embedding-001",
    client_args={"proxy": "socks5://user:pass@host:port"},
)

METHOD	DESCRIPTION
`embed_documents`	Embed a list of strings.
`embed_query`	Embed a single text.
`aembed_documents`	Embed a list of strings asynchronously.
`aembed_query`	Embed a single text asynchronously.

client `class-attribute` `instance-attribute` ¶

client: Any = None

The Google GenAI client instance.

model `class-attribute` `instance-attribute` ¶

model: str = Field(...)

The name of the embedding model to use.

Example: 'gemini-embedding-001'

task_type `class-attribute` `instance-attribute` ¶

task_type: str | None = Field(default=None)

The task type.

Valid options include:

'TASK_TYPE_UNSPECIFIED'
'RETRIEVAL_QUERY'
'RETRIEVAL_DOCUMENT'
'SEMANTIC_SIMILARITY'
'CLASSIFICATION'
'CLUSTERING'
'QUESTION_ANSWERING'
'FACT_VERIFICATION'
'CODE_RETRIEVAL_QUERY'

See TaskType for details.

google_api_key `class-attribute` `instance-attribute` ¶

google_api_key: SecretStr | None = Field(
    alias="api_key",
    default_factory=secret_from_env(["GOOGLE_API_KEY", "GEMINI_API_KEY"], default=None),
)

The Google API key to use.

If not provided, will check the env vars GOOGLE_API_KEY and GEMINI_API_KEY.

credentials `class-attribute` `instance-attribute` ¶

credentials: Any = Field(default=None, exclude=True)

Custom credentials for Vertex AI authentication.

When provided, forces Vertex AI backend.

Accepts a google.auth.credentials.Credentials object.

vertexai `class-attribute` `instance-attribute` ¶

vertexai: bool | None = Field(default=None)

Whether to use Vertex AI backend.

If None (default), backend is automatically determined:

If GOOGLE_GENAI_USE_VERTEXAI env var is set, uses that value
If credentials parameter is provided, uses Vertex AI
If project parameter is provided, uses Vertex AI
Otherwise, uses Gemini Developer API

project `class-attribute` `instance-attribute` ¶

project: str | None = Field(default=None)

Google Cloud project ID (Vertex AI only).

Falls back to GOOGLE_CLOUD_PROJECT env var if not provided.

location `class-attribute` `instance-attribute` ¶

location: str | None = Field(
    default_factory=from_env("GOOGLE_CLOUD_LOCATION", default=None)
)

Google Cloud region (Vertex AI only).

Defaults to GOOGLE_CLOUD_LOCATION env var, then 'us-central1'.

base_url `class-attribute` `instance-attribute` ¶

base_url: str | None = Field(default=None)

The base URL to use for the API client.

additional_headers `class-attribute` `instance-attribute` ¶

additional_headers: dict[str, str] | None = Field(default=None)

Additional HTTP headers to include in API requests.

client_args `class-attribute` `instance-attribute` ¶

client_args: dict[str, Any] | None = Field(default=None)

Additional arguments to pass to the underlying HTTP client.

Applied to both sync and async clients.

request_options `class-attribute` `instance-attribute` ¶

request_options: dict | None = Field(default=None)

A dictionary of request options to pass to the Google API client.

Example: {'timeout': 10}

output_dimensionality `class-attribute` `instance-attribute` ¶

output_dimensionality: int | None = Field(default=None)

Default output dimensionality for embeddings.

If set, all embed calls use this dimension unless explicitly overridden.

embed_documents ¶

embed_documents(
    texts: list[str],
    *,
    batch_size: int = _DEFAULT_BATCH_SIZE,
    task_type: str | None = None,
    titles: list[str] | None = None,
    output_dimensionality: int | None = None,
) -> list[list[float]]

Embed a list of strings.

Google Generative AI currently sets a max batch size of 100 strings.

PARAMETER	DESCRIPTION
`texts`	The list of strings to embed. TYPE: `list[str]`
`batch_size`	Batch size of embeddings to send to the model TYPE: `int` DEFAULT: `_DEFAULT_BATCH_SIZE`
`task_type`	`task_type` TYPE: `str \| None` DEFAULT: `None`
`titles`	Optional list of titles for texts provided. Only applicable when `TaskType` is `'RETRIEVAL_DOCUMENT'`. TYPE: `list[str] \| None` DEFAULT: `None`
`output_dimensionality`	Optional reduced dimension for the output embedding. TYPE: `int \| None` DEFAULT: `None`

RETURNS	DESCRIPTION
`list[list[float]]`	List of embeddings, one for each text.

embed_query ¶

embed_query(
    text: str,
    *,
    task_type: str | None = None,
    title: str | None = None,
    output_dimensionality: int | None = None,
) -> list[float]

Embed a single text.

PARAMETER	DESCRIPTION
`text`	The text to embed. TYPE: `str`
`task_type`	`task_type` TYPE: `str \| None` DEFAULT: `None`
`title`	Optional title for the text. Only applicable when `TaskType` is `'RETRIEVAL_DOCUMENT'`. TYPE: `str \| None` DEFAULT: `None`
`output_dimensionality`	Optional reduced dimension for the output embedding. TYPE: `int \| None` DEFAULT: `None`

RETURNS	DESCRIPTION
`list[float]`	Embedding for the text.

aembed_documents `async` ¶

aembed_documents(
    texts: list[str],
    *,
    batch_size: int = _DEFAULT_BATCH_SIZE,
    task_type: str | None = None,
    titles: list[str] | None = None,
    output_dimensionality: int | None = None,
) -> list[list[float]]

Embed a list of strings asynchronously.

Google Generative AI currently sets a max batch size of 100 strings.

PARAMETER	DESCRIPTION
`texts`	The list of strings to embed. TYPE: `list[str]`
`batch_size`	The batch size of embeddings to send to the model TYPE: `int` DEFAULT: `_DEFAULT_BATCH_SIZE`
`task_type`	`task_type` TYPE: `str \| None` DEFAULT: `None`
`titles`	Optional list of titles for texts provided. Only applicable when `TaskType` is `'RETRIEVAL_DOCUMENT'`. TYPE: `list[str] \| None` DEFAULT: `None`
`output_dimensionality`	Optional reduced dimension for the output embedding. TYPE: `int \| None` DEFAULT: `None`

RETURNS	DESCRIPTION
`list[list[float]]`	List of embeddings, one for each text.

aembed_query `async` ¶

aembed_query(
    text: str,
    *,
    task_type: str | None = None,
    title: str | None = None,
    output_dimensionality: int | None = None,
) -> list[float]

Embed a single text asynchronously.

PARAMETER	DESCRIPTION
`text`	The text to embed. TYPE: `str`
`task_type`	`task_type` TYPE: `str \| None` DEFAULT: `None`
`title`	Optional title for the text. Only applicable when `TaskType` is `'RETRIEVAL_DOCUMENT'`. TYPE: `str \| None` DEFAULT: `None`
`output_dimensionality`	Optional reduced dimension for the output embedding. TYPE: `int \| None` DEFAULT: `None`

RETURNS	DESCRIPTION
`list[float]`	Embedding for the text.

GoogleGenerativeAIEmbeddings¶