Anthropic announced on Thursday that it has updated the voice mode of its Claude assistant to allow users to choose between the Opus, Sonnet and Haiku models. The change is intended to make spoken interactions better suited for longer and more complex conversations.
Claude’s voice mode was launched last year and was originally powered by the Haiku model; that initial setup delivered fast responses but was less optimal for complex tasks. Under the new update, the voice mode defaults to whichever model the user last selected in the text chat and uses that model’s fastest variant.
Use cases for the updated voice mode
Anthropic says the revised voice mode can assist with longer conversations and practical tasks such as:
- giving feedback on a user’s communication style,
- rehearsing a pitch to a client,
- brainstorming product‑market research ideas.
In addition, the voice mode can now connect to other applications: supported integrations include Gmail, Google Calendar, Slack, Canva and Notion. With these connections, users can ask the assistant by voice to update a meeting slot, draft an email, or create a document in Notion.
Anthropic notes that this capability marks a distinction from some competitors: while OpenAI recently updated ChatGPT’s voice mode to change conversational style, that offering does not enable the assistant to use different tools to complete work.
Multilingual support and access limits
Earlier this year Anthropic added multilingual support to Claude’s voice mode in beta. The company says users can now speak in various languages, but the language must be specified manually. Supported languages are English, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese (Brazilian), and Spanish (Latin America/Spain).
The new voice mode is available to all users in beta across platforms. However, free users are limited: they can only use the Haiku model and may connect just one app.
What the update does not change
Anthropic clarified that this release did not modify the underlying voice model, and the company has not disclosed details about the voice stack it uses. As a result, unlike some recent competitor releases, users should not expect inherent conversational improvements such as better interruption handling.
