Issues with OpenAI plugins

Ask us anything
1

I use the built-in OpenAI plugin.
Simple and standard settings.

A few weeks ago, I thought it was just network issues. Or that it was happening because I was in edit mode. Unfortunately, it only works one out of five times in live mode too…
It’s impossible to go into production with this.

Has anyone else had this issue?
I’m still using the free version. Would switching to a domain name fix the problem?

The error that occurs from time to time:

name: "TypeError"
stack: "TypeError: Failed to fetch at Object.createChatCompletion (https://cdn.weweb.io/components/d66a250d-8468-469e-ad33-ee028f632398/def3f40a-4029-4bfa-a498-45c9cd0e84fd/dist/manager.js?v=1:1:51237) ..."
message: "Failed to fetch"

screen
screen1669×445 121 KB

2

Hi Yan :waving_hand:

Sorry for the late reply, did you manage to find a solution for this?

Did you open a support chat to figure this out?

3

Hi Augustin,

It seems it depends on the user internet connection. When I use the wifi, the error shows up, while all the saas still working perfectly. When I share my phone connection, much fewer errors.

So, I wrote an Edged Function, working perfectly in postman, n8n and others, to avoid the Weweb OpenAI plugin. But it seems Weweb is not able to handle the stream feature coming from Supabase. I asked to the support giving all the context, but I guess was too complex maybe for a fast help.

Until now the best option I have is to go to production finger crossed

1 Like