Skip to main content

Choice of local models for STT

Feature Requests

Not having a choice in speech-to-text models for the dictation feature is very limiting. The two pre‑selected models are not great. Silences are always replaced with lots of random words. And the user doesn't have the ability to pick from a whole variety of excellent speech-to-text models now available. Look at something like Mac Whisper and the flows that exist there.

Posted by Thinkun ·
4 comments

Log in to comment and vote

Comments4

  • Inspired Earth

    •

    May 13

    I’m also interested to know what it is you believe xAI can do that’s better than the current options (OpenAI Whisper, local Parakeet, etc.). I find both of those options work pretty flawlessly, so I’m curious how (in your experience) xAI improves on that?

    BTW, I am also a long-term user of MacWhisper, which is a great app. These days I use Parakeet almost exclusively in MacW and find it pretty flawless when transcribing clear recordings.

    • Daniel Nguyen

      Team•

      May 15

      Yeah in MacWhisper, the Parakeet models are considered “Pro” tier. It’s the default in BoltAI. Maybe the root problem is something else, not the model. Can you elaborate more @Thinkun

  • David

    •

    May 12

    i would love to see xAI in STT part… currently its only in Read Aloud

  • Daniel Nguyen

    Team•

    May 12

    Which models do you think are better? Supporting each model is not just changing a config. It requires a lot of work. Many users request this feature originally because they tested and believe Parakeet is better. BoltAI does support cloud models too (via OpenAI / Groq providers)