🧬 Embeddings

Embedding Models

We currently support the following embeddings models in our platform:

• BAAI (opens in a new tab)

  • bge-base-en-v1.5 (768 dimensions)
  • bge-large-en-v1.5 (1024 dimensions)
  • bge-small-en-v1.5 (384 dimensions)

• OpenAI (opens in a new tab)

  • text-embedding-3-large (3072 dimensions)
    • Also support variable dimensionalities of 128, 256, 512, 1024 and 2048
  • text-embedding-3-small (1536 dimensions)
    • Also support variable dimensionalities of 128, 256, 512 and 1024
  • text-embedding-ada-002 (1536 dimensions)

• Voyage AI (opens in a new tab)

  • voyage-large-2 (1536 dimensions)
  • voyage-2 (1024 dimensions)

• Cohere (opens in a new tab)

  • embed-english-v3.0 (1024 dimensions)
  • embed-multilingual-v3.0 (1024 dimensions)
  • embed-english-light-v3.0 (384 dimensions)
  • embed-multilingual-light-v3.0 (384 dimensions)

• JinaAI (opens in a new tab)

  • jina-embeddings-v2-base-en (768 dimensions)
  • jina-embeddings-v2-small-en (512 dimensions)

• NomicAI (opens in a new tab)

  • nomic-embed-text-v1.5 (768 dimensions)
  • nomic-embed-text-v1 (768 dimensions)

• MistralAI (opens in a new tab)

  • mistral-embed (1024 dimensions)

• intfloat (opens in a new tab)

  • e5-large-v2 (1024 dimensions)
  • e5-base-v2 (768 dimensions)

• thenlper (opens in a new tab)

  • gte-large (1024 dimensions)
  • gte-base (768 dimensions)

• MixedbreadAI (opens in a new tab)

  • mxbai-embed-large-v1 (1024 dimensions)
  • mxbai-embed-2d-large-v1 (1024 dimensions)

For customized and fine-tuned embedding models for your specific domain and use case, please contact us (opens in a new tab).

Usage

To list all embedding models available in our platform:

results = client.list_embedding_models()

You could add third-party API keys for the embedding models in our Dashboard (opens in a new tab).