Google Embeddings

Google embedding components provide access to Google's advanced embedding models through Google AI and Vertex AI platforms.

Google Generative AI Embeddings

This component generates embeddings using Google's Generative AI models.

Usage

Google Generative AI embedding features:

• High-quality text embeddings

• Multi-language support

• Optimized for semantic understanding

• Fast inference times

• Production-ready scaling

Inputs

Name	Display Name	Info
model	Model	Google embedding model name
google_api_key	API Key	Google AI API key for authentication
task_type	Task Type	Embedding task type (e.g., retrieval_query, retrieval_document)

Outputs

Name	Type	Description
embeddings	Embeddings	Google Generative AI embeddings instance

VertexAI Embeddings

This component provides embeddings through Google Cloud's Vertex AI platform.

Usage

Vertex AI embedding capabilities:

• Enterprise-grade embeddings

• Google Cloud integration

• Advanced model options

• Batch processing support

• Fine-tuning capabilities

Inputs

Name	Display Name	Info
model_name	Model Name	Vertex AI embedding model
project	Project ID	Google Cloud project identifier
location	Location	Google Cloud region
credentials	Credentials	Service account credentials path

Outputs

Name	Type	Description
embeddings	Embeddings	Vertex AI embeddings instance

Model Options

Available Models

Google Generative AI Models:

• embedding-001: General-purpose embedding model

• text-embedding-004: Latest text embedding model

• text-multilingual-embedding-002: Multilingual embedding support

Vertex AI Models:

• textembedding-gecko: General text embeddings

• textembedding-gecko-multilingual: Multilingual support

• text-embedding-preview-0409: Preview embedding model

• text-multilingual-embedding-002: Advanced multilingual model

Task Types

• RETRIEVAL_QUERY: Optimize for search queries

• RETRIEVAL_DOCUMENT: Optimize for document indexing

• SEMANTIC_SIMILARITY: General semantic understanding

• CLASSIFICATION: Text classification tasks

• CLUSTERING: Document clustering applications

Usage Notes

• Authentication: Requires Google Cloud credentials or API keys

• Regional Availability: Check model availability in your region

• Rate Limits: Respect API quotas and rate limiting

• Batch Processing: Efficient batch embedding generation

• Multi-language: Comprehensive multilingual support

• Enterprise Integration: Seamless Google Cloud ecosystem integration