I see numerous questions, but none seem to address this issue:
On the pricing page, it says GPT-4 API (or playground) has 8k Context.
However when using the API I frequently get an error saying I exceeded the 4096 context length limit.
nothing special to share really. I am using a standard GPT-4 ChatCompletion API call, with max_tokens = 1024, and I get an error saying my messages plus completion size exceeds the max context length 4096 of the model.