Commercial Embeddings

Commercial embedding components provide access to advanced embedding models from leading AI companies and specialized providers.

Cohere Embeddings

This component generates embeddings using Cohere's advanced embedding models.

Usage

Cohere embedding features:

• Multilingual embeddings

• Task-specific optimization

• High-quality semantic understanding

• Enterprise-grade performance

• Flexible model options

Inputs

Name	Display Name	Info
model	Model	Cohere embedding model name
api_key	API Key	Cohere API key for authentication
input_type	Input Type	Type of input (search_document, search_query, classification, clustering)

Outputs

Name	Type	Description
embeddings	Embeddings	Cohere embeddings instance

MistralAI Embeddings

This component provides embeddings through Mistral AI's embedding models.

Usage

Mistral AI embedding capabilities:

• Efficient embedding models

• Multilingual support

• Cost-effective processing

• European AI compliance

• High-performance inference

Inputs

Name	Display Name	Info
model	Model	Mistral AI embedding model
api_key	API Key	Mistral AI API key
endpoint	Endpoint	Custom endpoint URL (optional)

Outputs

Name	Type	Description
embeddings	Embeddings	Mistral AI embeddings instance

NVIDIA Embeddings

This component generates embeddings using NVIDIA's AI models and infrastructure.

Usage

NVIDIA embedding features:

• GPU-accelerated processing

• High-performance models

• Enterprise deployment

• Advanced optimization

• Large-scale processing

Inputs

Name	Display Name	Info
model_name	Model Name	NVIDIA embedding model
api_key	API Key	NVIDIA API key
base_url	Base URL	NVIDIA API base URL

Outputs

Name	Type	Description
embeddings	Embeddings	NVIDIA embeddings instance

Available Models

Cohere Models

• embed-english-v3.0: Advanced English embeddings

• embed-multilingual-v3.0: Multilingual embedding support

• embed-english-light-v3.0: Lightweight English model

• embed-multilingual-light-v3.0: Lightweight multilingual model

Mistral AI Models

• mistral-embed: General-purpose embedding model

• mistral-embed-large: High-capacity embedding model

NVIDIA Models

• NV-Embed-QA: Question-answering optimized embeddings

• e5-large-v2: General-purpose large embeddings

• nvolveqa_40k: Extended context embeddings

Task Optimization

Input Types

• search_document: Optimize for document indexing

• search_query: Optimize for search queries

• classification: Text classification tasks

• clustering: Document clustering applications

Use Cases

• Semantic Search: High-quality similarity matching

• Document Retrieval: Efficient document finding

• Content Classification: Automated categorization

• Similarity Analysis: Content comparison and analysis

Usage Notes

• Performance: High-quality embeddings with fast inference

• Scalability: Enterprise-grade scaling capabilities

• Multilingual: Comprehensive language support

• Task Specific: Optimized for different use cases

• Cost Effective: Competitive pricing models

• API Integration: Simple REST API integration