Rate Limit Exceeded BUG stopping PAID USE


some of our customers are getting this message in response to openai API calls. we are NOT anywhere near the rate limit. This is calling GPT 4.

If we call 3.5 then we get the message below. Read that error message… This is not good.


Really disappointing to see no response from OpenAI here. This is real Enterprise software with Fortune 500 customers who cannot use your service. The error messages clearly show something wrong happening with your service. You are competing against software companies that address bug reports.

A rate limit of 10000 tokens per minute is good for one client at best unless you limit the inputs and the conversation. I suspect it is more to satisfy the “when do I get GPT4 to put API keys into personality site” late-comers instead of the “when can I offer fortune 500 services” people.

user: Rewrite this 5000 tokens of code
bot: OK:…
user: try again
bot: rate limit error