Embedding Models

Configure which embedding model RAGaaS uses to understand your content.

Overview

Embedding models convert text into vectors that capture semantic meaning, enabling search based on understanding rather than exact matches.

When using embedding models, these texts would be considered similar:

"How do I cancel my subscription?" ≈ "What's the process for ending my membership?"
"Getting database connection timeout" ≈ "Database connection failed: timeout error"
"What's the pricing for enterprise plan?" ≈ "How much does it cost for large companies?"

This semantic matching helps find relevant content even when the exact words don't match.

During Ingestion:
- Your content is split into chunks
- Each chunk is converted to a vector
- Vectors are stored in your vector database
During Search:
- Your search query is converted to a vector
- Similar vectors are found
- Most relevant matches are returned

In RAGaaS, embedding models are configured at the namespace level. You'll need to provide your own OpenAI or Cohere API key.

{
  "embeddingModelConfig": {
    "provider": "OPENAI",
    "model": "text-embedding-3-small",
    "apiKey": "your-openai-key"
  }
}

{
  "embeddingModelConfig": {
    "provider": "COHERE",
    "model": "embed-multilingual-v3.0",
    "apiKey": "your-cohere-key"
  }
}

{
  "embeddingModelConfig": {
    "provider": "COHERE",
    "model": "embed-english-light-v3.0",
    "apiKey": "your-cohere-key"
  }
}

{
  "embeddingModelConfig": {
    "provider": "JINA",
    "model": "jina-embeddings-v3",
    "apiKey": "your-jina-key"
  }
}