Anthropic upgrades Claude's voice mode with access to more powerful models
Anthropic has expanded Claude's voice mode, letting users talk to the more capable Sonnet and Opus models instead of only Haiku. The feature can now also draw on connected apps like Gmail and Slack.
Since 2025, Anthropic has offered a voice mode for its Claude chatbot, letting users speak their prompts instead of typing them. This past July, the company made significant improvements to the feature.
Previously, voice mode was limited to Anthropic's least powerful model, Haiku, a choice made to keep response latency low. Now, voice conversations can also run through the more capable Sonnet and Opus models. The feature can additionally pull context from apps and services users have connected to Claude, such as Gmail, Google Calendar and Slack — for instance, asking Claude to summarize recent emails.
How voice mode works
Voice mode is available on Claude's mobile apps for Android and iOS, the desktop app, and the web, though Anthropic says it performs best on phones. The feature relies on a "turn-based" system, meaning Claude listens to a full prompt before generating a response. This differs from rival ChatGPT's voice mode, which OpenAI recently updated to a "duplex" architecture that processes speech and generates replies simultaneously. Because of this turn-based design, Anthropic recommends asking multi-part questions one at a time rather than all together, to avoid the system mistaking a pause for the end of a prompt.
Language and voice options
As of July 2026, voice mode supports 14 languages, with some offering multiple dialects — including Russian, relevant to many Baltic users. English offers five distinct voice options, while other languages may offer fewer. Users can also adjust the speaking speed and choose between hands-free and push-to-talk recording modes.
Free versus paid access
Voice mode does not require a paid Pro or Max subscription, but free accounts are restricted to the Haiku and Sonnet models and a single connected third-party app. Accessing the more powerful Opus model requires a paid subscription. All voice conversations count toward a user's usage limits.
