Hello all - could not find a specific post on this.
I am considering making custom GPTs on the Pro subscription, assigning each a role for developing a solution. However, am aware you can only “@” cross-reference custom GPTs using the 4.0 model, which is less advanced at coding and logic than the o3 mini-high or o1.
Does anyone know if using a single o3 mini high / o1 model makes any difference vs using 4.0 with cross-referencing custom GPTs? I know custom GPTs are trained and have an “edge” but if the model is less advanced / using less compute power etc and less advanced reasoning, does it actually make a difference? Or is this just one of those trade offs?
Thank you - mods please remove if you find a specific answer to this.