Audio model

Soniox TTS v1

Soniox TTS v1 is a real-time text to speech model that generates native-speaker-quality speech in more than 60 languages. It produces hallucination-free output with no invented words, dropped content, or unexpected substitutions, and supports streaming generation, multiple built-in voices, and configurable audio formats and bitrates. The model has been superseded by tts-rt-v2 and is deprecated.