outputAudioTranscription NOT WORKING WHEN [Modality.AUDIO]

Gemini API
, ,
1

Is there any update or timeline regarding when theAudioTranscription with [Mod.AUDIO] will fixed? My project on the transcript to detect specific phrases that trigger actions. Could you provide any information on when this issue might be resolved?

const model = ‘gemini-2.5-fl-preview-native-audio-dialog’;

My current configuration is as follows:

    config: {
      speech: {
        languageCode 'en-US',
      },
      responseModalities:ality.AUDIO], 
      outputAudioTranscription: {
      },
    }
3

Hi @Dorofino
As of now, Google has not published an official public reference or changelog entry confirming a fix or timeline for the outputAudioTranscription issue with [Modality.AUDIO] in gemini-2.5-fl-preview-native-audio-dialog.
You can monitor updates here:
Google AI Studio Release Notes
Gemini API Documentation
Thanks!

5

Yeah it’s currently not working, but I found a work-around: