OpenAI Launches GPT-Live With Real-Time Voice Conversations For ChatGPT
OpenAI launched GPT-Live on July 8, 2026, a full-duplex speech model that lets ChatGPT listen and speak simultaneously for natural, continuous conversations.
July 9, 2026: OpenAI launched GPT-Live on July 8, 2026, a new family of voice models that enables ChatGPT to listen and speak simultaneously, replacing its previous turn-based voice system with a speech-native model designed to support more natural and responsive conversations. The rollout is underway across ChatGPT on iOS, Android, and the web, with Application Programming Interface (API) availability planned for developers in the coming weeks.
Unlike the previous voice mode, which relied on separate speech recognition, language generation, and text-to-speech models, GPT-Live uses a full-duplex speech model that continuously processes audio while generating spoken responses. According to OpenAI, users can interrupt naturally, pause mid-conversation, and continue speaking without waiting for the assistant to finish. The system can also delegate complex tasks, including web searches and advanced reasoning, to OpenAI’s latest language models while maintaining the conversation.
OpenAI is rolling out two versions of the technology. GPT-Live-1 is available to Go, Plus, and Pro subscribers, while GPT-Live-1 mini is the default voice model for free ChatGPT users. The company said the models also support capabilities such as live translation and contextual visual cards that display information including weather forecasts, stock prices, and sports scores during voice conversations.
GPT-Live builds on OpenAI’s broader voice strategy. The company first introduced voice conversations to ChatGPT in 2023, before expanding the experience with GPT-4o’s Advanced Voice Mode in 2024. GPT-Live replaces that earlier architecture with an integrated speech-to-speech model designed to deliver more fluid, continuous conversations.
According to OpenAI, more than 150 million people use ChatGPT’s voice and dictation features each month. During a media briefing, product lead Atty Eleti said the company believes voice could become “the future interface to all kinds of work” as AI systems evolve to manage increasingly complex and long-running tasks, according to TechCrunch.
The company also announced that GPT-Live will be available through the OpenAI API in the coming weeks, allowing developers to integrate the real-time voice models into their own applications. OpenAI said the models are designed for conversational use cases that require continuous, low-latency speech interactions.
The launch expands OpenAI’s portfolio of multimodal AI capabilities by introducing a speech-native model that combines listening and speaking in real time. For organizations building AI-powered applications, the upcoming API release provides access to the same voice technology that powers ChatGPT’s new conversational experience, while the rollout expands access to the feature across ChatGPT’s free and paid plans.
