we are getting a 429 on paid account API access for over 24 hours?
this makes it unuseable in production.
yes, we have implemented backoff, etc.
but 24 hours later we still are getting 429 issues.
any ideas?
we are getting a 429 on paid account API access for over 24 hours?
this makes it unuseable in production.
yes, we have implemented backoff, etc.
but 24 hours later we still are getting 429 issues.
any ideas?
Hi,
How many attempts per minute are you sending, and how long has it been since you added a payment method to your account?
OK, what I’m trying to work out is are you getting 429 because you’ve used your limits for the month.
What is your current spend since the last billing cycle?
INVOICE | STATUS | AMOUNT | CREATED | |
---|---|---|---|---|
6B56332B-0019 | Paid | $1,891.13 | Oct 1, 2023, 9:59 PM |
$1,891.13
BUT: somehow we got approved for a $10K spending limit, so i am adjusting our limits now, I htink this was the issue.
ok,so we updated our “hard limit” for this month, now we get this error message:
{
“error”: {
“message”: “You exceeded your current quota, please check your plan and billing details.”,
“type”: “insufficient_quota”,
“param”: null,
“code”: “insufficient_quota”
}
}
even though it syas our spending limit is $10K
any ideas?
If the 10K spend was recently added, that can take time to effect your account limits, do you know when the 10k was added?
I think it would be a good idea to get the ball rolling with a help.openai.com visit and use the support bot in the bottom right corner to give your contact details and the nature of the issue, mention the 10K limit and that you are being prevented from making any new calls.
I’m pretty sure this is just some lag with with updates to your account, but it’s not a bad idea to get things started just in case.
yeah, cool, will do. kind of hard to put this stuff into production when the API does this. This is just us doing development, imagine in production? LOL!
p.s. it says their response time is within a week.
Yes, hence the get the ball rolling part, there are lots of people trying to use a limited resource. In most cases I see on here like yours, it’s resolved within 48 hours, but belt and braces with the help bot.
“Plan and billing details” is purely being shut off for a hard limit or spending limit on a monthly account. You’d review your billing overview and your usage page and see that monthly usage is not at either limit.
Plot the usage graph by “cumulative” and see where you are at for the month.
The figures may be a bit delayed, usage still being added up to a day after, but the effect of this should be allowed overages, not premature limits. Hard limit setting of yours should take effect almost immediately, lifting the cap you imposed yourself, and you can reset the soft email limit higher too.
(Also, we see from the information offered the order of who gets billed on the first of the month, and who gets billed two weeks later… )
Check organizations, default organization for API keys, organization if you use that in API calls, finally even a new API key.
If what you see isn’t being respected, database problems etc, that’s nothing that anyone but OpenAI staff can fix.
ok, 100% weird.
Plot the usage graph by “cumulative” and see where you are at for the month.
if we have a $10K limit, not anywhere even close LO.
but we just did a test call, and 429 is gone.
thank you whoever fixed that.
Great news, glad to see you’re up and running again!
Nice. They could improve by adding something like, “Once you raise the limit, it will take XX minutes/hours to reflect in your calls…” or a close estimate?
yeah, isn’t there a chatbot for chatgpt openai? lol there should be.
I think eventually, yup, there will be. I think if GPT-4 had all of the forum posts as training and it was kept up to date in real time, 95% of what goes on here could be automated.