I’m actually curious to know how you’re hosting whisper Vs say just using the OpenAI api. I’m also wondering how offloading the transcription would to the client with something like https://github.com/xenova/whisper-web