Gpt-3.5-turbo extremely frequent timeouts

[cobygoldstein]

I’m having the same issue with GpT-3.5

“name”: “Timeout”,
“message”: “Request timed out: HTTPSConnectionPool(host=‘api.openai.com’, port=443): Read timed out. (read timeout=600)”,

[storywriter]

Same here, seeing about 5% of the requests from my application time out. Started yesterday for me.

[fangtq229]

same here for both gpt-3.5-turbo and text-davinci-003

[jayben71]

API is almost non usable for the last 2 hours??? OpenAI please look into this asap.

[khaledgabr]

Same here unfortunately. It’s especially bad today. I hope response times gets improved.

[ruby_coder]

Yeah it is strange performance is so bad but status is so good, haha

[senols]

The performance has been significantly slower since yesterday, and it appears that OpenAI may not be aware of the issue yet. Completing a request is currently taking around 5 minutes.

[tylerprahm]

I am also experiencing this issue too. My requests timeout inconsistently

[IcyTorpedo158]

"Something went wrong. If this issue persists please contact us through our help center at help.openai.com. "

I’m getting this for the last 1 hr, won’t even respond to my friendly “Hi” lol

Not sure what’s going on…

[sbreslin]

Have been having the same issue over the last 12 hours.

[moritz1]

Looks like the website is now down completely. I just get a white and empty page.

[jendrik]

Same here. Really annoying. Does anybody know a workaround or faster / working model?

[DaveC]

Same issue. Calls to API that took a few seconds three days ago now taking 5 minutes or more

[duerr.simon]

I noticed some issues with the chat API as well. Often times reponse time is decent, but other times it fails on timeout. Horrible for user experience of course

[dpiret]

https://api.openai.com/v1/chat/completions is taking between 2s and a 5-minute timeout

[bptstmlgt]

Can confirm this is happening a lot today. Really hard to imagine building a decent UX on top of this.

[RabbitHam]

Obligatory +1. I’ve implemented exponential backoff, but subsequent retries still hang. Often, my entire retry limit is exhausted with no response with the ChatCompletions API with gpt-3.5-turbo.

[maxpain]

We have exactly the same problem in our project.

[bptstmlgt]

Now on the status page. For me, it’s been on and off since a few days.

[robeatz20]

Still getting very frequent errors… Timeouts, 502s…