Anthropic brings voice mode to Opus and Sonnet, expands into apps and languages

Anthropic is significantly expanding its voice mode, previously limited to the faster Haiku model, by making it available on its more capable Opus and Sonnet models. The company also announced broader third‑party integration, including support for apps like Gmail, Slack, and Canva.

According to Anthropic, users quickly adopted voice mode for more than just quick questions—they were using it to work “through real business problems.” Haiku, designed for speed, “kept conversations quick, but not always deep.” The Opus and Sonnet models, built for complex reasoning, can now deliver richer responses and take actions on behalf of users—for example, turning a voice brainstorming session into a one‑page pitch or rescheduling calendar appointments when a train is late.

Voice mode now supports seamless switching between text and voice mid‑conversation, as well as toggling between models. A user might start a quick factual query on Haiku, then switch to Opus to explore the idea in depth without losing context.

Expanded language support

Anthropic also graduated voice mode from English‑only beta to full support for nine other languages: French, German, Spanish, Hindi, Indonesian, Italian, Japanese, Korean, and Portuguese.

The rollout brings voice interaction closer to a genuinely useful tool for developers and knowledge workers who need both speed and depth—especially when integrated directly into the apps they already use.