Embedding Model Configuration Guide

R2R aims to be provider agnostic and ships with built-in support for the following providers:

  • OpenAI
  • SentenceTransformers (local)

The R2R team is actively working to increase the scope of these offerings in time, in accordance to the priority in which new providers are requested.

Available Models

Anything supported by OpenAI, such as:


  • Use case: Suitable for general-purpose embedding tasks with efficient processing. Ideal for applications where speed and cost are critical.
  • Dimensions: 1536 - Indicates the size of the embedding vector.
  • Recommended batch size: 32 - Optimal number of items to process in a single request for balancing performance and throughput.
  • Pricing: Approximately 62,500 pages per dollar. High efficiency and cost-effective for large-scale applications.
  • More: Embeddings Guide (opens in a new tab)


  • Use case: Ideal for tasks requiring high-quality embeddings, such as semantic search or complex text similarity. Best for when the quality of the embedding is paramount.
  • Dimensions: 4096
  • Recommended batch size: 16
  • Pricing: Approximately 9,615 pages per dollar. Offers superior performance at a higher cost.
  • More: Embeddings Guide (opens in a new tab)


  • Use case: A balanced option for tasks needing a compromise between quality and efficiency. Suitable for a wide range of applications.
  • Dimensions: 2048
  • Recommended batch size: 24
  • Pricing: Approximately 12,500 pages per dollar. Balances cost and performance effectively.
  • More: Embeddings Guide (opens in a new tab)

Lastly, the sentence_transformer package from HuggingFace is also supported as a provider. For example, one such popular model is mixedbread-ai/mxbai-embed-large-v1 (opens in a new tab).