Anthropic has upgraded Claude's voice mode, now allowing paid users to interact with its more powerful Opus and Sonnet models through spoken conversations. Previously, all voice interactions were processed by Haiku, the smaller, faster Claude model, which was ideal for quick replies but struggled with complex tasks.
This update marks a significant shift. Voice mode now defaults to the last text chat model a user engaged with, ensuring faster and more context-aware exchanges. Users can also switch between Haiku, Opus, and Sonnet models mid-conversation via a new interface feature, bringing flexibility without interrupting the flow.
also the voice mode now integrates Claude’s app connectors. This enhancement enables spoken commands to access tools like Gmail, Google Calendar, Slack, Canva, and Notion. Users can therefore move meetings, compose emails, or create Notion documents just by speaking, greatly expanding practical use cases for work and productivity.
The improvements come shortly after OpenAI launched GPT-Live, but Anthropic emphasizes a different focus: prioritizing solid reasoning and tool integration over a purely speech-native experience. This strategic direction shows Anthropic’s intent to differentiate Claude in the competitive AI assistant landscape.
