ModelVoice models

GPT-Live-1

GPT-Live-1 is OpenAI's July 2026 ChatGPT Voice model for more natural spoken interaction. OpenAI says GPT-Live-1 and GPT-Live-1 mini can listen and speak at the same time, making turn-taking and interruptions feel more natural.

Why it matters

GPT-Live-1 is a clear signal that voice models are moving from turn-based chatbot speech toward full-duplex human-computer interaction. Readers need to know where it is available, what it can and cannot do, and why it is separate from API realtime voice models.

Source-backed summary

OpenAI Help Center release notes are the factual source for GPT-Live-1 in ChatGPT Voice: rollout date, paid/free model names, full-duplex behavior, ChatGPT surfaces, web search and memory behavior, visual widgets, image support, Business/Enterprise/Edu launch exclusion, and no video or screen sharing support at launch. X and media coverage only add demand and reader-language context.

Primary use cases

Understand the new ChatGPT Voice model and full-duplex behavior.
Compare ChatGPT Voice with API realtime voice models.
Plan voice UX around interruptions, memory, search, widgets, and image context.
Check whether Business, Enterprise, Edu, video, or screen-sharing requirements block adoption.

What OpenAI says changed

OpenAI says ChatGPT Voice is now powered by GPT-Live-1 for paid users and GPT-Live-1 mini for Free users. The key behavior is full-duplex conversation: the model can listen and speak at the same time, so interruptions and turn-taking feel less rigid than older voice modes.

Availability and boundaries

The release notes say GPT-Live-1 is rolling out on chatgpt.com and the ChatGPT iOS and Android apps in supported regions. It is not available in ChatGPT Business, Enterprise, or Edu workspaces at launch, and it does not support video or screen sharing at launch.

Why it is entity-only for now

GPT-Live-1 has strong explainer value, but the current evidence is ChatGPT Voice availability rather than a stable standalone API catalog record for GetLLMs. Keep API realtime voice models, if needed, as a separate page family with official API docs and model IDs.

Related concepts

AI Model API

Use API model records only when OpenAI exposes stable standalone API fields.

Agent Harness

Voice agents need a harness around memory, search, visual results, and user interruption handling.

Related entities

GPT-5.6

OpenAI frontier model family that may be delegated to for complex text reasoning outside voice turn handling.

Sources

Source confidence

official-docs

ChatGPT release notes: Introducing GPT-Live-1 in ChatGPT Voice

OpenAI Help Center

official-social

OpenAI GPT-Live announcement on X

OpenAI on X

official-docs

Advancing voice intelligence with new models in the API

OpenAI

GPT-Live-1 FAQ

Page-level questions for GPT-Live-1.

Is GPT-Live-1 available through the OpenAI API?+

The July 8, 2026 ChatGPT release note describes GPT-Live-1 inside ChatGPT Voice. Do not assume it is a standalone API model until OpenAI API docs publish the model ID, endpoint, pricing, and limits. For API voice apps, check OpenAI's realtime voice model docs separately.

What does full-duplex voice mean for GPT-Live-1?+

Full-duplex voice means the model can listen and speak at the same time, making interruption handling and turn-taking feel more natural than a strict one-person-speaks-then-the-other-responds flow.

Does GPT-Live-1 support video or screen sharing?+

OpenAI says GPT-Live-1 does not support video or screen sharing at launch. Eligible subscribers can continue using those capabilities with Advanced Voice Mode where available.