Apparently it's based on GPT-3.5. I wonder how it compares to ChatGPT with GPT-4 for more advanced problems, for instance stuff which requires combining multiple web APIs or some degree of logic.
It seems like you're also only allowed to ask 5 questions per day, hopefully that limit will increase (or better yet, go away) once it leaves beta.
If it's just trying to tweak a specific example it will probably be quite useful.
It seems like you're also only allowed to ask 5 questions per day, hopefully that limit will increase (or better yet, go away) once it leaves beta.