OpenAI has expanded ChatGPT's premium voice mode to the web platform, allowing users to engage in voice conversations with the AI chatbot directly through their browsers. This update was announced by the company's Chief Product Officer, Kevin Weil.

This week, the feature became available to ChatGPT's paid subscribers, including Plus, Enterprise, Teams, and Edu users. Previously, the advanced voice mode was introduced in ChatGPT's iOS and Android applications in September.

The premium voice mode utilizes OpenAI's GPT-4 native audio capabilities to enable natural, real-time dialogues between users and ChatGPT. The chatbot can understand and respond to non-verbal cues, such as speech rate, and possesses the ability to deliver emotional responses.

To start a voice conversation on the web, users need to click the microphone icon located at the bottom right of the ChatGPT prompt window and grant the browser access to their computer's microphone. Afterward, users will be directed to a screen featuring a central blue sphere.

ChatGPT offers nine different voice output options, each with its unique tone and characteristics. For example, the "Arbor" voice is described as "amiable and versatile," while "Ember" is characterized by a "confident and optimistic" tone.

Kevin Weil stated that OpenAI plans to roll out this feature to free users within the "coming weeks."

Plus and Team subscribers will face daily time limits when using the premium voice mode. According to the feature's help page, these daily restrictions may vary. OpenAI will notify users when they have 15 minutes of premium voice usage remaining for the day. Free users will receive a monthly preview opportunity to try out this feature.