Anthropic has expanded its voice mode to run on the Opus and Sonnet models, moving beyond the previous restriction to the Haiku model. Users can now switch between these capabilities mid-conversation within the mobile app, desktop application, and web interface without restarting a session. The feature supports eleven languages and connects to external services such as Gmail, Google Calendar, and Slack to compose and send emails directly through audio input.
This expansion matters because it aligns the voice interface with the highest tier of reasoning available to the service. While competitors like OpenAI and Google offer full-duplex audio that feels more natural during conversation, Claude maintains a distinct advantage in practical utility. It remains the only provider allowing users to draft and save emails entirely from audio mode. This capability demonstrates how voice interfaces can execute complex tasks rather than just conducting simple queries.
* Voice mode now available on Opus and Sonnet models
* Switching between models occurs without interrupting conversation
* Direct email composition works via audio input
