“A query on a reasoning model like OpenAI's O1 could be up to 50 times more expensive than on a standard model due to increased token generation and reduced server utilization.”