When I tried to get transcription from Gemini Live API with Modality.AUDIO I am not seeing the transcription text in JSON response why?
config: {
responseModalities: [Modality.AUDIO],
outputAudioTranscription: {},
mediaResolution: MediaResolution.MEDIA_RESOLUTION_MEDIUM,
speechConfig: {
languageCode: 'en-US',
voiceConfig: { prebuiltVoiceConfig: { voiceName: 'Puck' } }
},
}
Did you try Modality.TEXT
?
Yes I have tried but I want both Audio and Text which not possible with Modality.TEXT
In your config place empty objects for output and input transcription:
const config: LiveConnectConfig = {
responseModalities: [ Modality.AUDIO ],
mediaResolution: MediaResolution.MEDIA_RESOLUTION_MEDIUM,
temperature: Number.parseFloat(assistantConfig.temperature),
inputAudioTranscription: {
},
outputAudioTranscription: {
},
...
HI, i’m using TypeScript and here is my config, why i still dont see the transcript?
config: {
speechConfig: {
languageCode: 'en-US',
},
responseModalities: [Modality.AUDIO],
outputAudioTranscription: {
},
},