OpenAI has introduced GPT-Realtime-2 alongside a series of new real-time voice models, aiming to position voice as the primary interface for AI interactions. These models are designed to facilitate more natural communication by enabling systems to listen, reason, and respond in real time. Available through the Realtime API, they provide developers with the tools to create voice agents capable of managing complex, context-rich conversations, supporting various applications such as live translation and guided workflows.

OpenAI: OpenAI is an artificial intelligence research and deployment company focused on developing advanced models for general intelligence applications. In May 2026, it released new realtime voice models in its API, including GPT-Realtime-2, designed to handle reasoning, translation, transcription, and action-taking during live conversations. This move positions the company to advance voice capabilities as a core part of its product ecosystem.

Product Launch: OpenAI introduced GPT-Realtime-2 and related realtime voice models to enable more natural voice interfaces that listen, reason, and respond in real time.
Developer Tools: New models are available through the Realtime API, allowing developers to build voice agents that handle complex, context-rich conversations.
Strategic Focus: The company is emphasizing voice as a primary interface for software interactions, supporting use cases like live translation and guided workflows.