I have FUTO voice input installed on my phone, so I have a local running whisper model running as the android voice recog default. it would be nice to be able to choose this because I can then use my own voice model without sending my voice to the cloud. Plus I can have custom settings in futo voice input for personalization (specific words and such)
The voice recog button in the kagi search widget already uses this android api to call the default voice recog app.