2h ago
Allow automatic voice language detection to be limited to selected languages
I would like to suggest an option to limit automatic language detection for voice input to a user-selected set of languages.
Sometimes the model incorrectly detects my speech as a language I don’t speak, even though I consistently use only two or three languages. It would be useful to keep automatic detection enabled, but allow users to specify which languages it should consider.
For example, I could select English, Ukrainian, and Russian, and the system would automatically detect between those three languages instead of considering dozens of other languages.
This would make automatic voice input much more predictable and reliable for multilingual users.
Pending