OpenAI launches new voice intelligence features in its API, including GPT-Realtime-2 with reasoning capabilities, real-time translation across more than 70 input languages, and live transcription.

OpenAI released three new voice models through its Realtime API: GPT-Realtime-2 (a voice model built with GPT-5-class reasoning for handling complex requests), GPT-Realtime-Translate (real-time translation supporting more than 70 input languages and 13 output languages), and GPT-Realtime-Whisper (live speech-to-text capabilities).

GPT-Realtime-2 differs from its predecessor GPT-Realtime-1.5 by incorporating GPT-5-class reasoning. Translate and Whisper are billed by the minute, while GPT-Realtime-2 is billed by token consumption.

OpenAI built guardrails into the system so that conversations can be halted if detected as violating its harmful content guidelines, designed to prevent misuse for spam, fraud, or online abuse. The company notes the features are intended for customer service, education, media, events, and creator platforms.