Overview
- OpenAI rolled out the voice-based agent features on ChatGPT’s web and mobile apps Wednesday to let people start multi-step workflows by speaking.
- The update links GPT‑Live — OpenAI’s simultaneous listen-and-speak audio model — to agentic models and a cloud browser so the assistant can browse, use plugins, and act across apps.
- Plus and Pro subscribers gain expanded Work-tab voice automation for documents, emails, Slack summaries, presentations, site building, and finance tools while Free and Go users keep plugin and connected-app access.
- OpenAI demonstrated real-world voice-only tasks, including analyzing spending, booking a $5 pickleball court, and reorganizing calendars, and said more than 150 million people already use ChatGPT’s dictation or voice modes.
- The move is pitched as both an immediate hands-free productivity boost and a step toward broader personal-assistant and possible hardware plans such as a reported smart speaker for 2027.