At a Glance · AI-generated

OpenAI released three new realtime audio models through its API on May 7-8, including GPT-Realtime-2, GPT-Realtime-Translate, and GPT-Realtime-Whisper. The models bring GPT-5-level reasoning capabilities to real-time voice conversations, enabling developers to integrate advanced voice intelligence features into their applications. The launch expands OpenAI's API offerings with enhanced audio processing and translation functionality for interactive voice interactions.