Claude's voice mode just got smarter
Anthropic has upgraded Claude's voice mode to use more powerful models and integrate with apps like Gmail and Slack, though some limitations remain.
Since last year, Anthropic has offered a voice mode for its AI assistant Claude, allowing users to speak instead of type. Previously, voice mode only used the Haiku model, Anthropic's smallest and fastest but weakest for complex tasks. Now, the company has released an update that lets voice mode tap into the more powerful Sonnet and Opus models.
For paying users, the voice mode will default to the last model used in text chat. Users can switch between Haiku, Sonnet, or Opus mid-conversation via the model picker. Anthropic notes that voice mode uses the fastest version of the selected model to keep the conversation smooth.
Additionally, voice mode can now pull context from connected apps like Gmail and Slack, provided the user grants permission. However, the mode is turn-based: Claude listens, pauses to think, and then responds. It is not fully duplex like OpenAI's GPT-Live system, which can process speech and generate output simultaneously. This makes talking to Claude potentially less natural than ChatGPT.
Another limitation is that Claude cannot automatically detect language switches mid-conversation. Users must either announce the switch or select the new language from the voice settings menu. On the upside, Anthropic has added support for more languages, including Indonesian.
The updated voice mode is rolling out in beta to all users across desktop, mobile, and web clients. Free accounts are limited to one connection and all queries go through Haiku, though they can speak in all supported languages.
