Hello Tim,
Where I add OpenAI API 1 and 2, would it be possible to add fields for “effort” and a fallback model?
I need the “effort” option because I use an endpoint based on OpenAI templates, but with Chinese models like GLM and DeepSeek; the issue is that if the effort level is too high, the model “thinks” too much, causing an SCM timeout with no response.
If the effort and fallback fields are filled in, they would be used; otherwise, nothing happens. I’m getting a lot of call errors due to this issue with the Chinese models overthinking things; if I set the effort to “low,” they proceed straight to writing the article without any problems.
This improvement applies to the “Page Pipeline” article creation section used for bulk article generation.
I saw that you suggested something along those lines at Fallback to different AI models and logins, but I believe it’s time to implement it.
What do you think?
