OpenAI API Introduces Realtime Voice Translation Supporting 70 Input Languages

OpenAI expands its API with realtime voice-to-voice translation, enabling speech input in over 70 languages translated into 13 output languages for operational use.

OpenAIAPITranslationVoice AIEnterprise AI

At a glance

OpenAI has added realtime voice-to-voice translation capabilities to its API. The feature supports speech input in more than 70 languages and output translation into 13 languages.

What changed

The OpenAI API now includes a realtime translation endpoint that processes spoken input across 70+ languages and delivers translated speech in one of 13 supported output languages. The update enables direct voice-to-voice workflows without intermediate text steps.

Why it matters

Teams can reduce manual transcription and translation steps, lowering workflow time and operational cost in multilingual support, research, and customer operations. Commercially, the capability allows providers to deliver faster localized services and expand addressable markets. For compliance-aware teams the feature supports consistent, auditable translation processes in regulated cross-border communications.

Key details

Input languages: 70+
Output languages: 13
Delivery: realtime voice-to-voice
Integration: available via OpenAI API

Sources

Notes for citation

This article is based on the official announcement shared by OpenAI leadership on X dated 29 May 2026. Operators should validate current language lists and API rate limits directly in the OpenAI platform documentation before production deployment.

Want to discuss how this affects your workflows? Book a call →

AI-assisted analysis by Skirr AI