I’ve built an AI agent in n8n and it works well most of the time, but I’m running into occasional request timeout issues when using GPT-5. Looking for advice on how to prevent or handle these timeouts more reliably. Any guidance or best practices would be appreciated.
If you go to options in the ai agent node you can set a batch size, which will lighten the load at once. There isn’t a max timeout option, but set the batch size to 1 and let me know how that works!
srry i misinputed didnt meant to send this post.
It looks like he is using the n8n cloud version - on my self hosted community edition, i can change the timeout for the OpenAI Chat node, but it seems you cannot add that parameter in the cloud version.
P.S.: i just found that on my community edition, the OpenAI node has a default timeout of 60 Seconds, whereas the OpenRouter node has a default timeout of 360 Seconds - i don’t know if this is also the case for the cloud version, but maybe it could be a workaround to set up an OpenRouter Account and access ChatGPT via OpenRouter. Does anybody have any experience with this?
3 Likes
I use cloud and self host, but thats a very valuable find. You could theoretically setup a webhook server on the self host, process the data on the self host, and send it back to cloud and you could change the timeout on self host.
Yes, I am on a cloud version and how can i move to self hosted?
There a several providers that offer one-click installation of n8n on a virtual server, or you could install it on your own hardware using this guide by n8n:
You should be aware though, that you have to take care of updates and security issues yourself.