OpenAI's new voice mode arrives on ChatGPT desktop
OpenAI's new voice mode is now in the ChatGPT desktop app, letting users speak to control AI agents and run computer tasks. Powered by ChatGPT-Live models launched earlier this month, ChatGPT Voice works with ChatGPT Work and Codex and can handle multi-step commands that the phone version could not execute.
Key Takeaways
- ChatGPT Voice is rolling out globally in the ChatGPT desktop app.
- It works with ChatGPT Work and Codex to direct agents by voice.
- The feature uses OpenAI's ChatGPT-Live (GPT-Live) voice models.
- On macOS, Appshots can share on-screen context, including alt-text.
- Users can also tap Codex Voice via the iOS app through remote access.
According to TechCrunch, OpenAI said Thursday it has updated the desktop client so people can talk to the app to control AI agents and perform tasks on their computer. For readers tracking similar launches, see more coverage in our Future Tech & AI Wonders hub.
What can ChatGPT Voice do on desktop?
On desktop, ChatGPT Voice is built to take action, not just chat. OpenAI said users can dictate complex, multi-step commands and reply when ChatGPT needs more input during a session.
In a demo video, a developer asked ChatGPT in one spoken command to create a new thread, make a pull request, and find the root cause of a bug. The company also said the mode can use computer-use skills to look up websites and apps while coordinating work.
How does OpenAI's new voice mode differ from mobile?
When ChatGPT Voice launched on smartphones, it offered smoother conversations and better interruption handling, but it was not built to take action on phones. Thursday's desktop update is more capable for real workflows that span several steps.
OpenAI said the experience is powered by GPT-Live, so it can speak, listen, and coordinate work in the app at the same time. That matters for users who want hands-free control while agents keep running in ChatGPT Work or Codex.
Does it work with ChatGPT Work, Codex, and macOS?
Yes. OpenAI said ChatGPT Voice works with both ChatGPT Work and Codex, letting users direct multiple agents by voice. On macOS, Appshots can let the app access what is on screen, including alt-text for extra context.
The company also said users can utilize ChatGPT Voice in Codex from the iOS app through remote access. Anthropic has likewise updated Claude's voice mode to complete tasks in apps such as Gmail, Calendar, Slack, Notion, and Canva, underscoring a wider shift toward voice-driven assistants.