OpenAI has unveiled GPT-Live, a new generation of voice models designed to power real-time, natural conversations on ChatGPT Voice. The system allows the AI to listen and respond simultaneously, a departure from traditional systems that wait for a user to finish speaking. This capability enables the model to handle instant interruptions, recognize pauses, and provide brief conversational cues like “yeah” or “mhmm” to maintain a human-like flow.

Integration and Performance

GPT-Live integrates directly with OpenAI’s broader ecosystem. While standard voice interactions are handled natively, the model automatically routes complex requests—such as web searches or advanced reasoning—to a separate frontier model before delivering the results through the audio interface. The system will initially rely on GPT-5.5 for heavier workloads and will update as newer models become available.

Internal testing conducted by the company showed improvements over its previous Advanced Voice Mode. Metrics where gains were recorded include conversational flow, interruption handling, and user preference during conversations lasting five to 10 minutes.

Feature Expansion and Safety

The update expands ChatGPT Voice beyond audio responses. The feature will begin displaying visual cards during voice conversations for topics such as weather, stocks, and sports. Access to tools including search, memory, image generation, and file uploads will be maintained.

New safety measures include protections against unauthorized voice imitation. The system uses a limited set of approved voices rather than allowing users to create direct replicas of real people. The company acknowledged that performance may vary across languages, with some languages experiencing differences in accent accuracy and fluency. OpenAI stated it has optimized GPT-Live for popular languages and is actively working to improve the experience where gaps exist.

Competitive Landscape

The release marks a step in the race among AI developers to move beyond text-based chatbots toward persistent, conversational assistants. Google has been expanding its Gemini-powered voice capabilities, while Amazon is rebuilding Alexa around generative AI. Apple is working to integrate advanced AI features into Siri through its Apple Intelligence platform. Startups like ElevenLabs are also advancing in AI-generated speech and conversational applications.

OpenAI has introduced GPT-Lite, launching two versions—GPT-Lite and GPT-Lite Mini—to ChatGPT users globally. The rollout begins today, marking an expansion of the platform's offerings.

Model Details

The new models include:

The introduction of these models aims to enhance user experience by providing varied options for different use cases.