FlashNEWS BRIEFING

OpenAI Opens GPT-Live-1.1 Voice Model API

1 min read

88Z AI, 20 November — OpenAI has released the GPT-Live-1.1 voice model API for developers, enabling full-duplex real-time bidirectional conversations. The model detects pauses and interruptions, improves multilingual and alphanumeric recognition, and delegates complex tasks to backend models for seamless agent experiences.

At a glance

  • Supports full-duplex real-time dialogue with pause and interruption detection
  • Delegates complex tasks to backend models for seamless multitasking
Cover image for OpenAI Opens GPT-Live-1.1 Voice Model API
Cover image for OpenAI Opens GPT-Live-1.1 Voice Model API

The update

OpenAI has officially opened the GPT-Live-1.1 voice model API to developers, aiming to provide full-duplex real-time bidirectional conversation capabilities that allow users to listen and respond simultaneously. The model features significant optimizations in interaction details, accurately recognizing pauses, interruptions, and changes in speech speed. Furthermore, GPT-Live-1.1 improves multilingual processing and enhances the recognition accuracy of mixed numeric and alphabetic content. In terms of functional architecture, it supports delegating complex reasoning and tool-calling tasks to backend models, ensuring that voice conversations remain uninterrupted. This design enables an agent experience where chatting and task execution occur concurrently, greatly improving interaction coherence and efficiency.

Sources & editorial note

This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.

  1. IT之家Source published

Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.

Related reading

All stories
88Z.ai