Grok Bot is deploying a voice interface to its desktop and mobile applications beginning Sept. 17. The feature provides text-to-speech capabilities in 30+ voices across 25+ languages and will reach all users over the next 2 days. This update replaces text-only interaction with a low-latency audio system for the AI assistant.

The underlying Grok Voice model previously launched on the fal developer platform for building intelligent agents. This version of the tool answers in 0.70 seconds, completes tool calls before a sentence ends, and supports voice cloning from 2 minutes of audio. SpaceXAI designed the system to resolve real customer issues through these low-latency interactions.