Custom GPT's only use the turbo version. (To save money and resources apparently)

GPT builders
,
1

Just spent the evening creating custom gpts only to now learn that they run on 4 turbo or maybe now 4o turbo.

I am realy disappointed as over time I found the base 4 version provides much better quality, more accurate and relevant results, it also adheres to custom instructions much better.

If I want to use the base 4 model with a custom gpt I have to use API’s and all that which let me decide which model it uses, great. But I’d have to pay more money to use it! I already pay the monthly membership fee so this really grated on me.

So looks like I’m back to using the default chat with custom instructions and the past evening has been a total waste of time.

Does anyone know if they have plans to offer the choice of what model to use with custom gpts?

2

I did not see any plan till now for future about GPTs. Based on this page they already switched to GPT-4o.

3

So now if they are going to or have already changed it to 4o, I will have also lost what I liked most about the standard 4 model; its superior complex reasoning and logical consistency.

What is your experience with the 4o model? Are you using it with custom gpts? Am I missing something or worrying about nothing? It’s just I see many comments of people saying its gone backwards compared to base 4 model.

Thanks for the reply and nice to see you again polepole :slight_smile: