A: Correct, the ability to select your own AI model was removed after receiving feedback from all of our users that a) they couldn't tell a difference between the models so it was pretty much just a distraction, b) we got rid of credit based plans b/c people didn't like them so we are not able to offer premium models under these new plans that allow unlimited regenerations, and c) I've committed to making the AI model choice a responsibility of Video To Blog based on feedback from our users who say they just want "the best one".

We do a lot of testing between the models and yes, I agree sonnet 3.5 has a reputation for being the best with writing, but we actually did not find that to be the case with Video To Blog. I'm not exactly sure why that is (I assume because our writing is based on a transcript and not just creative writing coming out of thin air and/or some combination of the prompt we use), but yeah, myself (and all of users) could not see a difference with Sonnet 3.5 over GPT-4o-mini so it didn't justify the cost.

We are currently using GPT-4o-mini. As more models come out we will be testing them thoroughly to see if any significant improvement is made. I'm committed to always using the AI model that produces the highest quality content for our users.

All of that said, we have found the highest improvement in quality of our blog output comes from adding custom instructions to tell the AI exactly what you want.