OpenAI has launched GPT-Live, a new generation of voice models powering ChatGPT Voice, designed to make conversations faster, less interruptive, and more natural by listening and speaking at the same time.
TL;DR
- OpenAI is rolling out GPT-Live-1 and GPT-Live-1 mini globally for ChatGPT Voice.
- The full-duplex models can listen, speak, pause, and respond continuously.
- Paid users get GPT-Live-1, while Free users get GPT-Live-1 mini.
- OpenAI says the update also brings smarter answers, visual responses, and voice-specific safeguards.
OpenAI has launched GPT-Live, a new family of voice models built to make talking to ChatGPT feel closer to having a real conversation. The launch began on July 8, 2026, with GPT-Live-1 and GPT-Live-1 mini rolling out to ChatGPT users globally, according to OpenAI and Reuters.
The biggest change is the move to a full-duplex architecture. In simpler terms, GPT-Live can listen and speak at the same time, rather than waiting for a clearly defined turn to end. That means users can interrupt, pause, ask it to slow down, or ask it to stay quiet while they gather their thoughts.
This is a notable shift from older voice systems. OpenAI said the original ChatGPT Voice relied on speech-to-text, a language model, and text-to-speech systems, while Advanced Voice Mode reduced latency but still depended on turns and silence detection. GPT-Live, by contrast, continuously processes input while generating output, allowing it to decide whether to speak, listen, pause, interrupt, or use a tool.
During a briefing cited by The Verge, OpenAI research lead Kundan Kumar called GPT-Live-1 the company’s “smartest voice model” yet. OpenAI product lead Atty Eleti also explained, “it can speak and listen at the same time,” highlighting the central upgrade behind the new model.
The model also handles deeper tasks differently. When users ask questions that require web search, reasoning, or more complex work, GPT-Live can delegate those tasks to GPT-5.5 in the background, then bring the result back into the conversation. At launch, GPT-Live-1 and GPT-Live-1 mini use GPT-5.5 Instant in the background, while Medium and High reasoning options use GPT-5.5 Thinking.
Topics For More Insights
OpenAI is also positioning the upgrade as a broader ChatGPT Voice refresh. Each week, more than 150 million people use ChatGPT Voice and Dictation, according to the company. The new experience can acknowledge users with phrases such as “mhmm” or “got it,” deliver better listening in noisy environments, support live translation, and show visual cards for weather, stocks, sports, and more.
Availability follows a tiered rollout. GPT-Live is rolling out across iOS, Android, and ChatGPT.com. GPT-Live-1 will be the default for Go, Plus, and Pro users, while GPT-Live-1 mini will be the default for Free users. OpenAI’s Help Center adds that Live is not available in ChatGPT Business, Enterprise, or Edu workspaces at launch, and does not initially support video, screen sharing, connected apps, or plugins.
OpenAI has also emphasized safety. Its GPT-Live system card says the models include voice-specific safeguards, with checks on inputs and generated outputs as conversations unfold. The company says the system can steer or interrupt unsafe responses, play spoken safety messages, provide support resources in text, or end higher-risk voice conversations.
OpenAI now says GPT-Live will come to the API soon, though it has not provided a firm launch date. For users, the change is simple: ChatGPT Voice should now feel more responsive, more patient, and more capable.
