GeminiEmbedding

​

Parameters

Parameter	Type	Default	Description
api_key	Optional[str]	None	Google AI API key
model_name	str	"gemini-embedding-001"	Gemini embedding model
enable_safety_filtering	bool	True	Enable Google’s safety filtering
safety_settings	Dict[str, str]	{"HARM_CATEGORY_HARASSMENT": "BLOCK_MEDIUM_AND_ABOVE", "HARM_CATEGORY_HATE_SPEECH": "BLOCK_MEDIUM_AND_ABOVE", "HARM_CATEGORY_SEXUALLY_EXPLICIT": "BLOCK_MEDIUM_AND_ABOVE", "HARM_CATEGORY_DANGEROUS_CONTENT": "BLOCK_MEDIUM_AND_ABOVE"}	Safety filtering settings
task_type	str	"SEMANTIC_SIMILARITY"	Embedding task type
title	Optional[str]	None	Optional title for context
enable_batch_processing	bool	True	Enable batch processing optimization
use_google_cloud_auth	bool	False	Use Google Cloud authentication
project_id	Optional[str]	None	Google Cloud project ID
location	str	"us-central1"	Google Cloud location
requests_per_minute	int	60	Requests per minute limit
use_vertex_ai	bool	False	Use Vertex AI API instead of Gemini Developer API
api_version	str	"v1beta"	API version to use (v1beta, v1, v1alpha)
enable_caching	bool	False	Enable response caching
cache_ttl_seconds	int	3600	Cache TTL in seconds
output_dimensionality	Optional[int]	None	Output embedding dimension (128-3072)
embedding_config	Optional[Dict[str, Any]]	None	Additional embedding configuration

​

Functions

​

__init__

Initialize the GeminiEmbedding provider. Parameters:

• config (Optional[GeminiEmbeddingConfig]): Configuration object
• **kwargs: Additional configuration options

​

_setup_authentication

Setup Google authentication.

​

_setup_client

Setup Gemini client with new google-genai library.

​

_setup_caching

Setup response caching.

​

supported_modes

Get supported embedding modes. Returns:

• List[EmbeddingMode]: List of supported embedding modes

​

pricing_info

Get Google Gemini embedding pricing. Returns:

• Dict[str, float]: Pricing information

​

get_model_info

Get information about the current Gemini model. Returns:

• Dict[str, Any]: Model information

​

_get_task_type_for_mode

Map embedding mode to Gemini task type. Parameters:

• mode (EmbeddingMode): Embedding mode

Returns:

• str: Task type for the mode

​

_check_rate_limits

Check and enforce rate limits.

​

_embed_batch

Embed a batch of texts using Google Gemini with new API. Parameters:

• texts (List[str]): List of text strings to embed
• mode (EmbeddingMode): Embedding mode for optimization

Returns:

• List[List[float]]: List of embedding vectors

​

_embed_single_text

Embed a single text using Gemini with new API. Parameters:

• text (str): Text to embed
• task_type (str): Task type for embedding

Returns:

• List[float]: Embedding vector

​

_embed_texts_batch

Embed multiple texts in batch using new API. Parameters:

• texts (List[str]): List of texts to embed
• task_type (str): Task type for embedding

Returns:

• List[List[float]]: List of embedding vectors

​

validate_connection

Validate Gemini connection and model access. Returns:

• bool: True if connection is valid

​

get_usage_stats

Get detailed usage statistics. Returns:

• Dict[str, Any]: Usage statistics

​

get_safety_info

Get content safety and filtering information. Returns:

• Dict[str, Any]: Safety information

​

list_available_models

List available Gemini embedding models. Returns:

• List[Dict[str, Any]]: List of available models

​

close

Clean up Gemini client and clear resources.

​

create_gemini_document_embedding

Create Gemini embedding optimized for documents. Parameters:

• api_key (Optional[str]): Google AI API key
• **kwargs: Additional configuration options

Returns:

• GeminiEmbedding: Configured GeminiEmbedding instance

​

create_gemini_query_embedding

Create Gemini embedding optimized for queries. Parameters:

• api_key (Optional[str]): Google AI API key
• **kwargs: Additional configuration options

Returns:

• GeminiEmbedding: Configured GeminiEmbedding instance

​

create_gemini_semantic_embedding

Create Gemini embedding for semantic similarity. Parameters:

• api_key (Optional[str]): Google AI API key
• **kwargs: Additional configuration options

Returns:

• GeminiEmbedding: Configured GeminiEmbedding instance

​

create_gemini_cloud_embedding

Create Gemini embedding with Google Cloud authentication. Parameters:

• project_id (str): Google Cloud project ID
• location (str): Google Cloud location
• **kwargs: Additional configuration options

Returns:

• GeminiEmbedding: Configured GeminiEmbedding instance

​

create_gemini_vertex_embedding

Create Gemini embedding with Vertex AI. Parameters:

• project_id (str): Google Cloud project ID
• location (str): Google Cloud location
• **kwargs: Additional configuration options

Returns:

• GeminiEmbedding: Configured GeminiEmbedding instance