Block has unveiled a new voice generation model that the company claims can produce speech with a high degree of human-like realism. A notable feature of this technology is its ability to run locally on a MacBook M4 Max, rather than requiring cloud-based processing or external server support.

By enabling local execution, the model potentially offers users greater privacy and lower latency, as data does not need to be transmitted to a remote server to generate audio. While the announcement highlights the technical capability of the model to mimic human vocal patterns, it does not detail specific commercial applications or the underlying training data used to achieve this level of realism. The company’s focus on hardware compatibility suggests a shift toward making advanced generative AI tools accessible on standard high-end personal computers.