Clarification on Specifying Embedding Models in OpenAI Vector Stores

API
, ,
1

I’m seeking clarification regarding the specification of embedding models when creating vector stores using the OpenAI API.

According to the official documentation for the POST /v1/vector-stores endpoint, there is no mention of an embedding_model parameter. This raises the question: how does one specify a particular embedding model, such as text-embedding-3-small, during the creation of a vector store?

Could anyone confirm if this is the correct approach? Additionally, are there any plans to allow users to specify the embedding model directly when creating vector stores via the API?

https://platform.openai.com/docs/api-reference/vector-stores/create

curl https://api.openai.com/v1/vector_stores \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -H "OpenAI-Beta: assistants=v2" \
  -d '{
    "name": "Support FAQ"
  }'

Thank you for your assistance!

1 Like