Pricing clarification on embedding models
Aris Alder
0
Reputation points
I understand that if I run a 10k token document through text-search-curie-doc-001
that I will be charged on 10k tokens to embed.
However, when using text-search-curie-query-001
, if I run a 50 token prompt/query, will I only be charged 50 tokens?
How does interacting with the 10k token document impact pricing? If the "similar" text found in the 10k token document is 500 tokens and the query is 50 tokens, is the charge 550 tokens?
Please expand if I am missing any key points in my logic.
Sign in to answer