A few weeks after OpenAI launched a new generation of conversation models and upgraded the ChatGPT voice function, competitor Anthropic also took action and released a major update to the voice mode of its artificial intelligence assistant Claude. According to the adjustment announced by the company on Thursday (July 23, 2026), users can now freely switch between three different levels of models: Opus, Sonnet and Haiku.
Claude's voice mode was first launched last year. It was previously driven by the faster but limited Haiku model by default. Although it responded quickly, it was not suitable for handling complex tasks. Anthropic said that in this update, the speech mode will automatically match the fastest corresponding version by default based on the model last used by the user in the text chat.
This upgrade significantly enhances the voice assistant’s performance during long conversations. Officials pointed out that the new version of the voice mode can not only provide targeted feedback on users’ communication styles, but also assist users in practicing customer promotions and even jointly conceive and brainstorm product market research plans.
In addition to the improvement of core dialogue capabilities, another highlight of Claude’s voice mode is its ability to integrate the application ecosystem. The system can directly call third-party applications such as Gmail, Google Calendar, Slack, Canva, and Notion. This means users can update meeting schedules, draft emails, or create new documents in Notion directly through voice commands. In contrast, although OpenAI has updated the tone and expression style of speech, it is still unable to directly call external tools to complete actual work collaboratively.
In terms of language support, Anthropic has previously added multi-language functionality to the beta version. Currently, users need to manually switch and specify the desired language. Supported languages include English, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese (Brazil) and Spanish (Latin America/Spain).
Currently, the new voice mode is open to beta test users on all platforms. However, free users will be subject to certain limitations in experience. They can only use the Haiku model and can only connect to one external application at most. It is worth mentioning that Anthropic did not change the underlying voice architecture in this update, nor did it disclose the specific voice technology stack, which means that compared with OpenAI’s update, there may not be substantial improvements in the natural conversation experience such as interruption processing.