Released API released Sep 10, 2026
OpenAI / GPT-Live-1
GPT-Live-1 is available for voice apps and phone conversations OpenAI’s voice model can listen and speak simultaneously, handle interruptions, and delegate reasoning or actions to a separate backend model.
Audio OpenAI API
By Privaj Scout · Brief published Sep 11, 2026 · Sources checked Sep 16, 2026
SCOUT’S TAKE
It supplies the conversational voice layer. Budget for the reasoning model and other services your application uses as well.
Voice-layer price $0.05 per minute
Interaction Simultaneous listening and speaking
Phone calls Telephony supported
Custom voices Contact sales for eligibility What changed. 01 The API became available on September 10.
02 Developers can pair it with an OpenAI or third-party text model.
03 Custom voices require an eligibility and access request.
OpenAI API launch and pricing ↗ Before you use it. The published per-minute rate covers the voice layer. It is not an all-inclusive application price; backend-model and infrastructure costs depend on your setup.
What the evidence shows Availability, features, and launch pricing checked against OpenAI’s announcement on September 16. No independent latency result is claimed.
BEHIND THE BRIEF
Go to the source. Specifications and availability come from the maker. Scout’s take is editorial interpretation.
Sources checked Sep 16, 2026. Availability and requirements can change.
Read our methodology Suggest a correction Read another briefChatGPT Images 2.5 
ut