OpenAI Newsroom, AlphaSignal, AIHOT featured · Published 2026-09-10
GPT-Live-1: interruptible real-time voice for apps
GPT-Live-1 listens and speaks in the same live conversation, while a backend model can handle reasoning and tools.
Speaking does not require strict turns
OpenAI describes a full-duplex model that listens while speaking and responds to pauses, additions and interruptions. Conversation can continue while work happens in the background.
Voice and backend have different roles
The voice layer handles conversation; paired models can take on deeper reasoning and tool tasks. Developers choose that backend. The two pairs of arrows show speech exchange and delegated work returning a result.
The same pattern extends to phone calls
The announcement lists restaurant reservations and customer support as use cases. A caller adding details mid-conversation illustrates the interaction; completing a booking still depends on the connected systems, tools and permissions.
Sources and notes
sources中文
Sources and notes
- Read the originalOpenAI Newsroom, AlphaSignal, AIHOT featured
- GPT-Live guide: voice and backend delegation
Retained summary and context
Related sources may rely on the same material. Claims remain attributed to their sources. Collection is not proof of the claims or practical outcomes.
Sources and evidence · 3 links
Each link has a purpose and scope. Link counts do not establish independent confirmation; direct support applies only to the statements specified below.
This page keeps the original link and a short account. Follow the links below for the publisher’s full explanation.
Source timeline
- · Found this reading link via OpenAI Newsroom, AlphaSignal, AIHOT featured.
Collection record
First discovered: ; added to the library:
Collection and source publication have distinct dates.