I added my OpenAI key and URL. Anything I ask the AI results with a pink “Error — Not Found” error box. I can’t see any other details. Is there a way to turn on debug mode or something?
i have the same error after some minutes of conversion with chatgpt 4.1 mini
Daniel Nguyen
Team•
May 2
Sorry I missed this issue. Does it work now? I’ve since added quite a few more checks to prevent this.
terheyden
•
Dec 4, 2025
Never mind, I figured it out. The custom URL I specified was incorrect. I was using:
https://us.api.openai.com - instead of:
https://us.api.openai.com/v1
You can close this!
Daniel Nguyen
Team•
May 2
Sorry I missed this issue. I wasn’t aware that OpenAI allow setting custom regional based endpoint like that. Does it work now? I’ve since added quite a few more checks to prevent this, especially during provider setup. It should show more user friendly error now.
terheyden
•
Dec 4, 2025
Also, sorry, I tested with gpt-5.1 and gpt-5-mini models, same error message.
terheyden
•
Dec 4, 2025
BTW the same key and URL work perfectly fine in other CLI tools — I just tested.
Log in to comment and vote
Comments6
elbephant GmbH
Dec 19, 2025
i have the same error after some minutes of conversion with chatgpt 4.1 mini
Daniel Nguyen
May 2
Sorry I missed this issue. Does it work now? I’ve since added quite a few more checks to prevent this.
terheyden
Dec 4, 2025
Never mind, I figured it out. The custom URL I specified was incorrect. I was using:
https://us.api.openai.com - instead of:
https://us.api.openai.com/v1
You can close this!
Daniel Nguyen
May 2
Sorry I missed this issue. I wasn’t aware that OpenAI allow setting custom regional based endpoint like that. Does it work now? I’ve since added quite a few more checks to prevent this, especially during provider setup. It should show more user friendly error now.
terheyden
Dec 4, 2025
Also, sorry, I tested with gpt-5.1 and gpt-5-mini models, same error message.
terheyden
Dec 4, 2025
BTW the same key and URL work perfectly fine in other CLI tools — I just tested.