Jun 25
Allow per-model setting for auto-insert behavior after voice input
Could you please allow configuring whether the transcription output is automatically inserted after recording on a per-custom-model basis? At the moment, this setting seems to apply only at the global level.
Personally, I sometimes use voice input simply to pass prompts to something like GPT-4 and get a response, without needing the transcribed text inserted automatically.
In those situations—especially compared to coding environments—having the result injected into the text area can actually get in the way.
It would be very helpful if this behavior could be made optional per model. Thanks for considering it!
Pending
similar to https://superwhisper.userjot.com/board/p/separate-settings-p…