OpenAI has enhanced its ChatGPT desktop application by integrating ChatGPT Voice, enabling users to engage with AI agents and manage computer tasks through spoken commands.
This functionality is powered by OpenAI’s recently developed ChatGPT-Live voice model series, introduced earlier this month, which supports real-time interaction and processing.
The update expands compatibility across ChatGPT Work and Codex platforms, allowing voice-activated operations such as web browsing and app interactions. On macOS, the Appshots feature grants the AI access to the user’s screen contents, including alt-text descriptions.
Unlike the smartphone version, which focuses on conversational flow but lacks action-oriented capabilities, the desktop iteration supports complex, multi-step voice commands and interactive responses when additional input is required.
A demonstration video from OpenAI illustrates the technology’s potential: a developer instructing ChatGPT to create a new thread, initiate a pull request, and diagnose a bug—all through a single voice command.
ChatGPT Voice is now in the desktop app.
Control your computer and direct multiple agents running in ChatGPT Work or Codex, using just your voice.
It’s powered by GPT-Live, so it can speak, listen, and coordinate work in the app simultaneously.
Rolling out globally today… pic.twitter.com/ODZWKqecCf
— OpenAI (@OpenAI) July 23, 2026
Users can also access ChatGPT Voice features within the iOS Codex app via remote access capabilities.
In parallel, Anthropic has enhanced its Claude voice mode with expanded model options—Opus, Sonnet, and Haiku—to support task automation across applications like Gmail, Slack, and Notion.
