@OpenAI: Audio moves through a dedicated fast path, while deeper reasoning and tool use happen asynchronously. We also reduced v…
Summary
OpenAI announces that audio now uses a dedicated fast path while reasoning and tool use run asynchronously, and voice-session startup has been reduced from six network round trips to one.
View Cached Full Text
Cached at: 08/03/26, 09:53 PM
Audio moves through a dedicated fast path, while deeper reasoning and tool use happen asynchronously.
We also reduced voice-session startup from six network round trips to one. https://t.co/XI8jvuxrDP
Similar Articles
How OpenAI Delivers Low-Latency Voice AI for 900M Users (17 minute read)
OpenAI's voice AI architecture uses WebRTC with a split relay-transceiver design to handle low-latency audio for 900 million weekly active users.
How OpenAI delivers low-latency voice AI at scale
OpenAI details its rearchitected WebRTC stack designed to deliver low-latency voice AI at scale for over 900 million users. The post explains how new split-relay and transceiver architectures optimize media routing and connection setup for real-time interactions like ChatGPT voice.
OpenAI's New Voice Models Want to Do More Than Talk Back
OpenAI has launched three new real-time audio models to enable continuous, multitasking voice interactions that prioritize long-context reasoning, live translation, and seamless tool use.
OPEN AI: How we built a realtime system for responsive voice AI in six months
OpenAI describes how they built GPT-Live, a full-duplex realtime voice AI system that eliminates the turn detector, enabling natural continuous conversation. The article details architecture improvements in inference, context management, and media transport over six months.
@oliviscusAI: NVIDIA just removed the biggest friction point in voice AI. They open-sourced PersonaPlex 7B, a real-time conversationa…
NVIDIA open-sourced PersonaPlex 7B, a real-time conversational model that listens and speaks simultaneously, handling natural interruptions and overlaps unlike most voice models.