The new model addresses programming issues and a UTF bug I was unaware of before.
To the cursor team - please test thoroughly and release it.
As my manager often says at work - I’d prefer you take an extra few days to perfect it than try to cross the finish line early and wreck it.
Fair to say that it’s not ideal to get stuck with gpt4 while gpt4 turbo keep getting updates?
Agreed, for the first 500 ‘fast requests’ there should at least be the option to choose, especially sinds Turbo should be cheaper. After that, using their in-house GPT-4 model that saves costs is fine by me.
Additionally, when can we except 0125-Turbo to be available through the API?
For those who are antsy to give the new model a try, you can now try out
gpt-4-0125-preview as a pro or API key user if you configure the model in settings. Please let me know if you run into any issues.
anyone have any thoughts/comments on the performance of