Friday, Nov 13 | --:--
Times of AI

Regions

China Europe India Japan Taiwan United Kingdom United States

Platforms

Alibaba Qwen Amazon Anthropic DeepMind Google Groq Meta Microsoft MiniMax Mistral Moonshot AI NVIDIA OpenAI PrismML Replit Thinking Machines xAI

Sections

Business Engineering Enterprise Frontier Labs Hardware Industry Infrastructure Models Policy Products Research Safety Security Work All articles → Subscribe by email
Back to home
Products Voice United States OpenAI

GPT-Live: Full-Duplex Voice First, Enterprise Later

OpenAI launched GPT-Live-1 and Live-1 mini for ChatGPT Voice—full-duplex listen-and-speak on consumer web/mobile; Business/Enterprise/Edu excluded at launch. Mass-user agent UX bid while governance surfaces wait.

Times of AI Desk 5 min read San Francisco, CA View as Markdown
ChatGPT Voice interface with dual waveform showing simultaneous listening and speaking under GPT-Live branding.

Voice is how mass users will live with agents. Full-duplex GPT-Live is OpenAI’s bid to make ChatGPT feel less like a walkie-talkie and more like a call—while keeping enterprise out of the first cut until safety and governance catch up.

OpenAI on July 8, 2026 launched GPT‑Live, voice models built for natural, full-duplex conversation—listening and speaking at the same time. GPT‑Live‑1 and GPT‑Live‑1 mini began rolling out globally in ChatGPT Voice (web, iOS, Android in supported regions). Paid consumer tiers get Live‑1; Free gets Live‑1 mini. At launch the stack used GPT‑5.5 in the background, with continuous upgrades planned as frontier models advance.

Product details#

From OpenAI’s “Introducing GPT‑Live” primary post, system card, and ChatGPT release notes:

  • Models: GPT‑Live‑1 and GPT‑Live‑1 mini.
  • Defaults: Live‑1 for Go/Plus/Pro; Live‑1 mini for Free.
  • Capability: Full-duplex—handle interruptions, pauses, and overlapping speech more naturally than turn-based voice modes.
  • Surfaces: ChatGPT consumer (chatgpt.com, iOS, Android); spoken responses with streamed text; web search, memory, widgets; text+image in the same conversation.
  • Not at launch: ChatGPT Business, Enterprise, Edu workspaces; no video/screen share on Live‑1 (Advanced Voice Mode remains for those capabilities where eligible).
  • Safety: Dedicated GPT-Live System Card published July 8.

Limits#

  • Business/Enterprise/Edu excluded at launch.
  • No video/screen share on Live‑1; Advanced Voice Mode remains for those where eligible.
  • Background reasoning model at launch was GPT‑5.5; upgrades planned as frontier models advance.
  • Naturalness claims are product framing; independent conversational MOS/latency benches were not published with the post.

Sources#

  • OpenAI: “Introducing GPT‑Live” (openai.com/index/introducing-gpt-live/, July 8, 2026). Primary.
  • OpenAI Deployment Safety Hub: GPT-Live System Card (July 8, 2026).
  • ChatGPT release notes: GPT-Live-1 / mini roll-out (July 8, 2026).
  • Reuters same-day coverage.

Related Coverage

Scroll to continue reading