On Thursday, OpenAI announced an upgrade to its ChatGPT desktop application that introduces ChatGPT Voice, enabling users to interact verbally with the AI and perform various tasks on their computers.
This new functionality leverages OpenAI’s latest voice model suite, known as ChatGPT-Live, which was recently unveiled.
According to OpenAI, ChatGPT Voice is compatible with both ChatGPT Work and Codex. It can also utilize skills related to computer usage to search for websites and applications. Additionally, on macOS, users can enable Appshots, allowing the app to access on-screen content, including alternative text descriptions.
When first launched, the mobile version of ChatGPT Voice offered seamless conversation flows with improved interruption handling, but it lacked the ability to perform actions on smartphones.
The latest desktop update is more powerful, enabling users to issue complex, multi-step commands and respond when prompted by ChatGPT for their input.
In a demonstration, OpenAI showcased a developer instructing ChatGPT to initiate a new thread, create a pull request, and identify the source of a bug using a single command.
Users can also use ChatGPT Voice with Codex through the iOS app by means of remote access.
Additionally, Anthropic has updated its voice capabilities for Claude, integrating it with the company’s Opus, Sonnet, and Haiku models for task completion in applications like Gmail, Calendar, Slack, Notion, and Canva.
