OpenAI Opens GPT-Live-1 Voice Model to Its API
Opening GPT-Live-1 to developers puts the full-duplex architecture behind third-party voice agents, including phone handling for reservations and customer support, not only ChatGPT's own voice mode.
Reporting from 1 source: GIGAZINE.
OpenAI announced an API for GPT-Live-1 on September 10, 2026. The voice model runs full-duplex, speaking while it listens, and can hand reasoning and tool calls to backend text models such as GPT-6 Astra or to third-party models. Developers set tone, speaking speed, and conversation style through system prompts. OpenAI claims a 30 percent gain over GPT-Realtime-2.1 on Full Duplex Bench.
The API lets developers tune tone, pace, and conversation style through system prompts, and lets the model pass heavy work such as web search and deep reasoning to backend text models while the conversation continues. OpenAI lists interruption handling, ambient noise and silence, and long-session context retention among the strengths it cites. Phone deployment is one named use, from reservations to customer support. A broader set of voice options spans accents, dialects, and languages, with more planned. OpenAI states GPT-Live-1 improved 30 percent over GPT-Realtime-2.1 on Full Duplex Bench.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.
Sources
- GIGAZINE OpenAIが音声会話AI「GPT-Live-1」のAPIをリリース