OpenAI's API pricing

I just wanted to know what window context was used for request with my own OpenAI API key for GPT-3.5 and GPT-4 to calculate the estimated costs.

Hey! If you use an API key, we’ll keep the context window under 8k. There’s not really any cost saving logic at the moment, if your current file, @'ed context, and conversation are 8k tokens, then we will fill out all 8k.

1 Like

and what is the context window for the pro subscription if i may know? @truell20

We largely rely on 8k but sometimes use 32k. Goal is to abstract this away and have things “just work.”

Hey hey hey, sorry for asking again but, is the windows context 4k for GPT-3.5 turbo or is it 16k?
Thanks in advance!


I think it’s very simple, they won’t cost you more than what you know, when you select gpt-3.5-turbo not gpt-3.5-turbo-16k (because there’s no option for it), then it’s using gpt-3.5-turbo, as 16k’s model costs more obviously.

1 Like

Okay thanks! Sorry if I bothered you or something, I just wanted to make sure :sweat_smile:

1 Like

You’re welcome, I’m sorry If I was harsh in my way :pray:

Amazing how the chatter in our heads affects how we each uniquely receive information? I guess it’s the human condition?

On another topic. Call me grateful, the Cursor team rocks!!!

1 Like