Just to clarify they might be accurate but only when using the absolute highest quality setting which is really really not necessary to get great results. The latest GPT image model has five quality levels and the first three are entirely sufficient for almost every use case
Just to clarify they might be accurate but only when using the absolute highest quality setting which is really really not necessary to get great results. The latest GPT image model has five quality levels and the first three are entirely sufficient for almost every use case