Grok TTS

Grok TTS is xAI's text-to-speech offering, designed to turn written input into spoken audio suitable for applications such as voice assistants, narration, and interactive agents. The model is referenced under the identifier xai.grok-tts and is positioned as part of xAI's broader Grok family, extending the line beyond text and image understanding into speech synthesis. Its purpose is to provide a flexible synthesis layer that developers can reach through standard API styles without needing to build bespoke audio pipelines.

In practical deployment, Grok TTS is exposed through two complementary pathways: an OpenAI-compatible Audio Speech API for straightforward request-and-response generation, and a WebSocket streaming endpoint that streams text in and audio out incrementally for lower-latency or interactive scenarios. The launch coverage points to five built-in voices and multilingual support spanning more than twenty languages, letting builders pick tone and accent without training separate models. Audio output can be tuned through standard parameters such as format, sample rate, and bit rate, making the model a practical fit for both batch content production and real-time conversational experiences.