To submit a feature request, please answer the following questions:
-
What problem does this feature request solve? - The OpenAI Whisper Large V3 model is the most accurate voice transcription model on the market at the moment. By adding this model to the voice-to-audio transcription options that you currently offer, will allow people to have access to the highest quality audio transcription in their workflows.
-
What is the use case for this feature? - I want to design a workflow that allows me to transcribe my podcast episodes and then use the text of the transcript in additional aspects of the workflow, such as: creating a summary, writing show descriptions, brainstorming title ideas, etc.
I currently do this offline on my computer using the large v3 model, but I’ve tested the transcription with several other models, including the ones you currently offer for voice transcription, and they are simply not as accurate as the v3 model. So I would love to have the v3 model available as an option.
-
Please describe the functionality of this feature request. - This should be relatively simple since you already offer the Whisper model in your settings. It would just be a matter of adding this particular version, V3.
-
Is there anything else we should know? - Nothing I can think of.
You can upvote on the features that you care about.