Skip to main content

Streaming Transcription

Coming Soon

Streaming transcription is not yet available. This page documents the planned API.

client.asr.transcribeStream() will provide real-time word-by-word output as audio is processed — ideal for live recordings, calls, or long files where you want results before processing completes.

Planned signature​

client.asr.transcribeStream(
  file: FileInput,
  options?: TranscribeStreamOptions,
  requestOptions?: RequestOptions
): AsyncIterable<TranscriptionStreamEvent>

Planned usage​

for await (const event of client.asr.transcribeStream(audioBuffer, { language: 'en' })) {
  if (event.type === 'chunk') {
    process.stdout.write(event.text);
  }
  if (event.type === 'complete') {
    console.log('\nJob ID:', event.jobId);
    console.log('Full transcript:', event.transcriptionText);
  }
}

Planned event types​

TranscriptionChunkEvent​

{
  type: 'chunk';
  text: string;
  startTime: number | undefined;
  endTime: number | undefined;
  confidence: number | undefined;
}

TranscriptionCompleteEvent​

{
  type: 'complete';
  jobId: string;
  transcriptionText: string;
  detectedLanguage: string | undefined;
  qualityScore: number | undefined;
  processingTimeMs: number | undefined;
}

In the meantime, use Batch Transcription for all transcription needs.