When using the chat-openai-com UI, it’s well understood that there is a usage cap of 40 messages / 3 hours on the GPT-4 model model=gpt-4.
However, when interacting with the Plugins model model=gpt-4-plugins, there is no such information presented and the model itself claims that it is not bound by a hourly limit.
(instance 1)
Qwerty
Are you gpt-3 or 4?
ChatGPT
I am based on GPT-4.
Qwerty
Do you have hourly limit?
ChatGPT
No, I don’t have an hourly usage limit. However, individual platforms or services using me might impose their own usage limits or restrictions.
However, just now, in another Plugins conversation, it says something different
(instance 2)
Qwerty
Do you have an hourly usage cap?
ChatGPT
Yes, I have an hourly usage cap. This is implemented to manage system load and ensure fair access for all users.
Yet the first instance still continues with its previous narrative. What gives?
(instance 1)
Qwerty
Do you have an hourly usage cap?
ChatGPT
No, I don’t have an hourly usage cap. However, the platform or service you’re using to access me might have its own usage policies or restrictions.
So, what is the actual limit for using this model?
And is the limit separate or shared with gpt-4 model?