OpenAI brings ChatGPT Voice to desktop app with computer control and AI agent coordination

OpenAI said on July 23 that ChatGPT Voice is now available in the desktop app, giving users a voice-first way to control their computers and manage multiple AI agents at the same time. The feature is powered by GPT-Live, OpenAI’s real-time voice model, which allows ChatGPT to listen, speak, and coordinate active tasks inside the app without requiring users to switch back and forth between typing and manual actions. According to OpenAI, the same voice interface can direct agents running in ChatGPT Work for office collaboration and in Codex for software development. The rollout started the same day on macOS and Windows and is being offered globally. Eligible plans include Plus, Pro, Business, Edu, and Enterprise. OpenAI also shared an approximately 69-second demo video alongside the announcement. Earlier, GPT-Live had appeared in a multitasking demo that showed it handling flight booking, weather checks, and itinerary syncing at the same time. The desktop integration now turns that capability into a single entry point for computer control and agent management.