OpenAI has introduced GPT-Realtime-2, GPT-Realtime-Translate, and GPT-Realtime-Whisper through its Realtime API, strengthening its push into real-time multilingual voice experiences.
The new models support 70+ input languages and 13 output languages, enabling more natural conversations, live transcription, and seamless voice interactions across global audiences. The technology is designed for use cases spanning customer service, media, education, and creator platforms, where instant multilingual communication is becoming increasingly valuable.
OpenAI has also embedded safeguards aimed at reducing harmful, deceptive, and fraudulent activity, reflecting the growing importance of trust and safety in voice AI systems.
The rollout signals a broader industry shift where voice AI is evolving beyond simple assistants into real-time communication infrastructure capable of powering global, AI-native interactions.Multilingual voice AI is becoming a foundational layer for global digital communication and customer engagement.
Bottom line: OpenAI’s latest Realtime API update highlights how voice, translation, and transcription are converging into scalable, AI-powered communication systems for enterprises and creators alike.
