What happens after the STT phase if we enable Open Transcription?

Here is a simple use case to illustrate my question:

  1. The user speaks the wake word and pauses;
  2. Rhasspy wakes up;
  3. The user then speaks the command;
  4. Rhasspy performs speach to text using Open Transcription;
  5. ?

The documentation states that rhasspy will " ignore any custom voice commands you’ve specified".


  1. What will happen then? Will it still try to do intent recognition?
  2. If it ignores any custom voice commands, how will it do intent recognition?

I’m also curious about this. I’m trying to look through documentation to figure it out but if anyone has answers / clues that would be great!

Well, I have just tried it out, and got even more confused. So the documentation states that it will ignore any custom voice commands. I was under the impression that means all the custom intents, and sentences. But this is not what is actually happening.

If it transcribes a sentence that is recognized by your intent manager then the intent will get recognized and it works as it always does.

If it transcibes something that is not recognized, than it will simply not recognize any intent. It works similar to the text input for me. If I type anything that my sentences do not reflect, i simply get an “error” message.