How long are your OpenAI instructions

Hello,

I’m still very much a beginner with APIs. I’m building an OpenAI call now where I’m inputting either text or an image and getting a JSON back. I’m finding I keep needing to enter additional rules/context into the prompt to get a better result. But is there a point where the prompt would get too long, making the call inefficient?

Each model has its context window…so it ultimately depends on the model you are using. Which model are you using?