OpenAI launches new voice intelligence features in its API

OpenAI launched new Realtime API models for voice apps: GPT-Realtime-2, GPT-Realtime-Translate, and GPT-Realtime-Whisper.

The update adds live translation in more than 70 input languages and 13 output languages, plus real-time transcription.

The models are built for talking, translating, and transcribing conversations in developer products.

Pricing is based on minutes or tokens, and OpenAI says guardrails are in place to limit harmful use.