Morning! Apologies if this has been asked before, I did look through the forums.
What is the realistic upper bound of requests per minute. I did ask the chatgpt and it said it can accomodate 100,000 requests per minute (did not specify token size), but what is the amount someone can drive through the API and do we know specifically what tier? Presumably Tier5?
Thanks in advance!
J
Welcome to the community, @John_Clark …
Have you checked out the rate limits page in the docs?
Hello! Appreciate the reply!
I did. I’m in tier 1, so I bumper up against our TDM, causing the API call to “fail”.
Is this independant of the tier? Presumably, if I pay more, I would be able to access the API, say up to 100,000 requests per minute.
Why I ask, I’m building out a team, and the team anticipates potential, absolute maximum of ~5,184,000 API invocations per day. Per minute, this breaks down to 3600 minute. In my current tier, this maxes me out. Do you know what tier I would have to be in, for this number of calls to be sufficient?
Perhaps I’m not being clear, I do appreciate your patience and working with me on this.
Hi,
Welcome to the forum.
You can find this information here:
https://platform.openai.com/settings/organization/limits
(I am Tier 4 and seem to have up to 10k Requests Per Minute Limit)
Just to note… Based on your usage TPM might end up being the bottleneck you should consider multiple axis above. (RPM, TPM, model, etc.)
