I see where I can regenerate a summary, but is there a way to regenerate a transcript?
I recorded something prior to having access to the Pro models, and the transcription was horrible. Is there a way to have a different AI model re-transcribe the audio?
Commenting more for the following aspect. I’m curious as well. I’d imagine that’s not how it works. But we’ll find out!
I haven’t tested this, but maybe try downloading or exporting the audio, have it play, and re-record and see what happens? 
That’s an option, but not one that makes Pocket more usable. Right now my main use for Pocket is for a 4 hour class every day. I don’t have a spare 4 hours to manually play/re-record.
Hey @DaveNgl, you can share your feature suggestion on our feedback board at https://feedback.heypocket.com/. We review all feature requests and prioritize them based on community interest.
Thanks for helping us improve Pocket!
Hey @bharat I left some feedback on there yesterday via comments and have received no reply. How monitored is that compared to this forum?
@Nex I’ll DM you to follow up directly.
I guess that’s a no? It’s not currently possible to regenerate a transcript?
@DaveNgl No, you cannot regen a transcript itself, only the Summary and Actionable based on the GPT and template you use
Pro subscription has nothing to do with transcription. The LLMs are only used for generating the summaries.
I understand that we can’t pick the LLM that does the transcribing, but one of them is used for the transcription.
hey @DaveNgl so LLM’s are not used for transcription. There are a different suite of models which do that for us internally. These are the best in the industry and have been finetuned for our usecases.
Interesting. Has Pocket used the same model the entire time?
we do iterate on the model internally to make it better every day! but yes we have used the same model from the start. we also do a lot of research internally to give you the best transcriptions possible