Grok will have a voice copying feature.

xAI is taking Grok's speech recognition capabilities to the next level with its upcoming Voice Transcription feature. This feature allows users to record their own voice by reading a specific piece of text, and Grok will then collect and process the recording to create a personalized voice profile.

 

The transcribed voice can then be used in Grok's existing voice mode or shared with others via a link. When someone opens a shared voice link, they also have the option to save it for their own use.

 

This takes xAI far beyond simply providing pre-set voices. The company has launched the Grok Voice Agent API, creating the fastest conversations on the market with response times under 700ms, and this API is now also used in Tesla vehicles. Voice cloning will be a further step, moving the product from pre-set voices to personalized audio.

Grok's voice sharing mechanism is particularly noteworthy. This link-based voice distribution model could pave the way for the formation and growth of an informal voice market, something some reports had predicted when platforms began deploying custom voice options in previous years.

This feature is still under development and has no official release date, although xAI's recent release pace for Grok suggests it's not far off.