Skip to main content

Claude Voice Mode Gets Smarter: Opus and Sonnet Join the Conversation

Anthropic has given Claude's voice mode a significant upgrade. Previously limited to the lightweight Haiku model for speed, the voice mode now supports all three Claude models: Opus, Sonnet, and Haiku. This means users can tap into the full reasoning power of Opus or Sonnet while speaking, turning voice interactions from simple Q&A into deep, real-time advisory sessions.

The update comes in response to how people actually use voice mode. Since Anthropic introduced hands-free conversation earlier this year, users quickly adopted it for more complex scenarios—long business discussions, brainstorming sessions, and strategic planning. But the Haiku-only limitation meant conversations were fast but often lacked depth. Now, with Opus and Sonnet available, users can choose the model that best fits their needs. A model selector lets you switch anytime during a conversation, and the voice mode defaults to the model you last used in text chats, so you can move between voice and text seamlessly.

Image

With stronger models, the voice experience changes fundamentally. Instead of getting standard answers, you can speak half-formed ideas and let Claude help you develop them. It asks questions, extends your thoughts, and challenges assumptions. Anthropic suggests several scenarios: practice a difficult conversation and get feedback on your communication style, organize a client proposal and have Claude identify logical gaps, brainstorm a product roadmap and ask for competitive research, or explore why a video went viral by testing different hypotheses.

Once you've refined your ideas, Claude can help execute them. If you're running late, ask it to reschedule a meeting on Google Calendar. Turn a client proposal discussion directly into a Canva document. Ask it to summarize today's emails and draft responses for the most important ones. Before using connected tools, Claude asks for permission. You can add new tools in Settings > Connectors on mobile, desktop, or web.

Language support has also expanded. Voice mode now works with more languages across all plans. You can speak in the language you think in, and switch languages mid-conversation. However, Claude doesn't automatically detect the language—you need to verbally request a switch or manually select it in voice settings. Also, voice mode language preferences are separate from text, so you'll need to set them independently.

The update is available in beta for all chat users on mobile, desktop, and web, though the experience is best on mobile. Free users get Haiku, one connected tool, and all supported languages. Paid plans unlock more models and all connected tools. Voice conversations count toward regular usage limits. To start, tap the sound wave icon in the Claude mobile app.

Key Points

  • Claude voice mode now supports Opus, Sonnet, and Haiku models, selectable mid-conversation.
  • Users can connect to tools like Gmail, Slack, and Canva for real-time task execution.
  • Language support expanded; users can switch languages during a conversation.
  • Free users get Haiku and one tool; paid users access all models and tools.
  • Best experience on mobile; available in beta on all platforms.