OpenAI Launches Production-Ready Realtime API for Voice Agents
The general availability release on August 29 introduces gpt-realtime, a more advanced speech-to-speech model with lower latency and new features.
OpenAI on August 29, 2025, officially moved its Realtime API out of beta and into general availability, launching a new, more advanced speech-to-speech model called gpt-realtime. The release is designed to empower developers and enterprises to build reliable, production-ready voice agents that can interact with users in a more natural and expressive way.
A More Capable Voice Model
Unlike traditional systems that chain together separate models for speech-to-text, processing, and text-to-speech, gpt-realtime processes and generates audio directly within a single model. This end-to-end architecture is designed to significantly reduce latency, preserve nuance, and produce more fluid voice interactions.
OpenAI …
Archive Access
This article is older than 24 hours. Create a free account to access our 7-day archive.