OpenAI Expands ChatGPT Voice Conversations to Web Browsers
OpenAI brings its Advanced Voice Mode to the web, allowing paying subscribers to have real-time audio conversations with ChatGPT directly in their browsers. The feature uses GPT-4o to understand non-verbal cues and respond with emotion.
OpenAI expands ChatGPT’s Advanced Voice Mode to the web, enabling users to talk directly to the AI chatbot through their browsers. Chief Product Officer Kevin Weil announces that the feature rolls out this week to paying customers, including Plus, Enterprise, Teams, and Edu subscribers. This web launch follows the initial debut of the voice feature on iOS and Android apps in September.
The Advanced Voice Mode leverages the native audio capabilities of GPT-4o to facilitate natural, real-time conversations. ChatGPT understands and responds to non-verbal cues such as the user's speaking speed and generates replies with appropriate emotion. Users start a conversation by clicking the Voice icon, granting microphone permissions, and choosing from nine distinct output voices that each offer a unique tone and character.
While the feature is currently limited to paid subscribers with a daily usage cap, OpenAI plans to roll it out to free users in the coming weeks. Free users receive a monthly preview to test the capability, while paid users get a notification when they have 15 minutes of advanced voice usage left for the day. OpenAI notes that these daily limits are subject to change over time.