Have you thought about adding another speech-to-text transcription provider besides OpenAI?

Like offering an ElevenLabs integration. They just released Scribe v2 this week—specifically the base model that runs out of the box with no extra mods.

Because when you test the base model—just standard transcription with no mods—and then compare it to a custom mod you’ve tuned, the difference is pretty noticeable: the base model mangles a lot of words and the transcription quality isn’t great.

Cost-wise, I think OpenAI models and ElevenLabs are more or less in the same price range, and the integration there is probably just as straightforward as OpenAI’s. So if it’s possible, it’d be great to add this as an option in the app so you can choose the base transcription model—meaning not a model for custom mods, but the base Scribe model for transcription, which then gets processed by the models set up in the mods.

Here’s a link to the new model from ElevenLabs:

https://elevenlabs.io/blog/introducing-scribe-v2