One frustration that I have is that Chat GPT automatically will change my LLM within a project, to 4o without me asking them to do it, nor will it let me change it back. It happens randomly.
When I create a project, I usually will upload a GPT 4.5 and a GPT o3 mini high LLM. Sometimes, if I need the capabilities of a 4o model, I’ll add one of those to the chat. What I don’t like, is when I’m conversing with an o3 mini high model, and then all of a sudden, I notice that it’s switched to a 4o model, and it won’t let me change it back. Why is that?
1 Like
Actually, I see now that after I asked it another question, it went back to o3 mini high, reasoned, and responded. That is really cool and possibly a glimpse of what GPT 5 will be like. The only thing I can add to this, is that it doesn’t change it reliably. E.g., before it switched, I had been asking it if it could read information within a spreadsheet located within the project. It told me no. I tried to upload the spreadsheet to it, and it still told me it couldn’t read it. So, I had a 4o model read it, summarize it, and then I copied and pasted that information into the o3 mini high model and asked it for deeper insight. As a result, it switched to 4o model and gave me 4o model insight. Then it created an intervention as 4o. Then, I asked it if it conforms with the goals for the client, and that is the time that it switched back to o3 mini high. Who knows… ?