Transcribe audio in real time with low latency
Extract voice profile signals such as emotion, vocal style, accent, age, and pitch
Stream audio bidirectionally over WebSocket
Provide synchronous transcription for complete audio files
Integrate with AI for adaptive and expressive responses
Configure voice, context, and turn-taking on a single model ID
Filter low-confidence voice profile results with configurable thresholds
Support transcription in over 30 languages