On Thursday, OpenAI said it had updated its ChatGPT desktop app to add support for ChatGPT Voice, allowing users to communicate with the app to control AI agents and perform tasks on their desktop.
The up-to-date feature leverages a up-to-date family of OpenAI voice models called ChatGPT-Live, which was released earlier this month.
OpenAI says ChatGPT Voice works with both ChatGPT Work and Codex, and can also exploit computer skills to search websites and applications. Additionally, on macOS, with Appshots, users can allow an app to access screen content, including alt text.
At launch, the smartphone version of ChatGPT Voice offered smoother conversations and better noise handling, but was not designed to take action on smartphones.
Thursday’s desktop update is more competent, allowing users to dictate elaborate commands involving multiple steps and respond when ChatGPT needs their input.
In a demo video, OpenAI showed a developer asking ChatGPT to create a up-to-date thread, submit a pull request, and find the root cause of the error with a single command.
The company said that users can exploit ChatGPT Voice on Codex from the iOS app via remote access.
Anthropic has also updated its voice mode for Claude, which can exploit Opus, Sonnet, and Haiku models to perform tasks in apps like Gmail, Calendar, Slack, Notion, and Canva.