Anthropic Upgrades Claude Voice Mode With Opus, Sonnet and Haiku Models

Anthropic has upgraded Claude’s voice mode with support for its Opus, Sonnet and Haiku models, expanding the AI assistant’s capabilities weeks after OpenAI introduced new conversational models and an updated ChatGPT voice experience.

Voice mode, first launched last year using the Haiku model, was designed for quick conversations but had limitations when handling more complex tasks. With the latest update, Claude automatically uses the last model selected in text chat and defaults to its fastest version for voice interactions.

According to Anthropic, the enhanced voice mode is better suited for longer conversations, helping users practice client presentations, receive communication feedback and brainstorm ideas for projects and market research.

Productivity app integrations

One of the biggest additions is support for connected apps, allowing Claude to work with services including Gmail, Google Calendar, Slack, Canva and Notion. Users can ask the assistant to draft emails, reschedule meetings or create documents without leaving a voice conversation.

This capability distinguishes Claude from OpenAI’s current voice mode, which offers more natural conversations but does not yet perform tasks across multiple connected productivity tools.

Available across all plans

Anthropic also expanded multilingual support introduced earlier this year. Users can speak in English, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Brazilian Portuguese and Spanish, although the desired language must still be selected manually.

The updated voice mode is rolling out in beta across all platforms. Free users can access the Haiku model and connect one external app, while paid subscribers can choose between Opus, Sonnet and Haiku with broader integrations.

Anthropic said the release does not introduce a new voice model or changes to its underlying voice technology, meaning conversational features such as interruption handling remain largely unchanged compared with OpenAI’s latest voice updates.