Status page error? API is still much slower than usual!

[jochenschultz]

Status page is wrong!

but that is actually wrong!!!

Status page error? API is still much slower than usual!

This is NOT resolved and should be classified as a CRITICAL OUTAGE.

Checkout the status page here:

OpenAI Status

[rgran001]

API is very very slow. Now 11:15am on east coast. Been having this issue for hours now.

[tion8888]

I’ve been experiencing extreme slowness with the threads and assistant apis since yesterday afternoon, Feb 24th, east coast. The issue seems context length dependent, the more messages in the thread, the slower to fetch messages and run assistant. run assistant t takes longer than 60 seconds consistently for longish contexts, but it used to only take 2 seconds max for the same contexts. this is effectively a stop work situation for many of my users since yesterday, and still ongoing.

[pb462162]

guess now is a good time for me to go get some coffee and a bagel lol

[jochenschultz]

I understand that reference haha

enjoy

[Kiki92130]

I checked and was told the issue had been resolved, but it appears to be down again.

[sandro.trabucco]

Yep, same in Italy very unstable and slow… since last 2 days. I guess is somewhat related with the rolling out of the free version of its Advanced Voice mode for chatgpt.

[info1158]

Very slow in Sweden as well. Hopefully gets better soon.

[tion8888]

It’s still unusably slow in New York as of 1:34pm ET. even after receiving the notification that the issue has been resolved. It is clearly not resolved. We appreciate that the OpenAI team is working hard to fix the issue. Please let us know if we can help test or provide any assistance.

[john2jairo]

Hi, I am having the same issue here in the UK it is extremely slow and seems to be cutting out as if it was a bad signal but it’s not.
When will this get resolved?

[niklasohrn139]

This is NOT resolved and should be classified as a CRITICAL OUTAGE.

[jochenschultz]

Thank you for reporting this. I see the same on my side. ChatGPT also is much slower than usual.

@PaulBellow could you look into that?

[Saira]

Same, for me Feb 24th, 25, 26

[jochenschultz]

just saying

        } catch (Exception
                | TimeOutException
                | ClientExceptionInterface
                | DecodingExceptionInterface
                | RedirectionExceptionInterface
                | ServerExceptionInterface
                | TransportExceptionInterface $e) {
            $this->logger->error(
                'Error in ChatUnifiedMessageServiceStrategy',
                ['exception' => $e]
            );
            $delayStamp = new DelayStamp($this->retryDelaySec * 1000);
            try {
                $payload = $unifiedMessage->getPayload();
                $payload->setAlternativeProvider($this->alternativeProvider);
                $unifiedMessage->setPayload($payload);
                $this->messageBus->dispatch($unifiedMessage, [$delayStamp]);

[jochenschultz]

On my side it got better now. Anyone still experiencing issues? Could you post which models you are using then?

[mikelmorrison70]

WTF is going on? ChatGPT4o api is stalling out, freezing up, and intermittently slowing to an unusable crawl. Any updates? Goes on any longer… should be getting compensated.

[tion8888]

The API is unusably slow for the third day in a row. This is unacceptable at this point. I’m more than happy to provide help testing and troubleshooting to those working in this issue at OpenAI. I can consistently reproduce > 60 second response times for calls that used to take milliseconds. New York, Thursday morning 9:34am.

[tion8888]

I’ve shared the model, assistant, thread ids and run ids with support. hoping OpenAI can take a look asap. This is still a stop work situation for my users and clearly others are still experiencing the extreme slowness.

Model: mostly gpt-4o-2024-05-13