HomeEverythingEducation
Equities & FundsCrypto & Digital AssetsAI & TechnologyBusiness & CorporateUS Politics & PolicyGeopolitics & Global RiskMacro, Rates & FXCommodities & EnergyEuropean Politics & MarketsAsia-PacificReal Estate & Property
← All Stories

OpenAI launches GPT-Live voice models for simultaneous listening and speaking

Created at 8 Jul · 5:56 PM2 sources↑ Market-relevant2 events
IN SHORT

OpenAI has released GPT-Live, a new family of real-time voice models that can listen and speak simultaneously, allowing for more natural conversations and interruptions. GPT-Live-1 mini will replace the current Advanced Voice Mode in ChatGPT for paid users.

✉Newsletter

PiQ Daily

Pick your topics. Get only what matters, on your cadence.

Key Numbers

30- to 40-minute-longconversations with voice feature
150 millionpeople talk to ChatGPT using voice features

Who's Involved

OpenAI
AI company that launched new voice models
Atty Eleti
Product lead for ChatGPT Voice
OpenAI launches GPT-Live voice models for simultaneous listening and speaking

↳ Why This Matters

The advancements in GPT-Live models aim to make AI interactions more seamless and intuitive, potentially transforming how users engage with computing for complex tasks and paving the way for voice to become a primary interface.

Key facts

  • OpenAI launched GPT-Live, a new family of voice models enabling simultaneous listening and speaking.
  • The models, GPT-Live-1 and GPT-Live-1 mini, aim for more natural conversations and better turn-taking.
  • GPT-Live-1 mini will become the default for ChatGPT's Advanced Voice Mode for paid users.
  • These full-duplex models allow for natural interruptions and features like live translation.
  • The models can leverage newer GPT models for advanced reasoning and task completion.
  • OpenAI sees voice as a potential primary interface for complex computing tasks.

OpenAI has released GPT-Live, a new suite of real-time voice models designed to enable simultaneous listening and speaking, aiming to create more natural and conversational AI interactions. The models, GPT-Live-1 and GPT-Live-1 mini, are now available globally.

These full-duplex models allow users to interrupt conversations naturally, a significant improvement over previous sequential processing methods. GPT-Live-1 mini will replace the existing Advanced Voice Mode in ChatGPT for paid users, while the larger GPT-Live-1 will also be accessible to them. The new models integrate with OpenAI's latest text models, such as GPT-5.5, to handle complex queries, reasoning, and agentic tasks while maintaining conversational flow.

During a press briefing, OpenAI demonstrated the models' ability to remain silent and absorb conversational context before responding, as well as their capacity to present information visually. The company believes that voice interfaces will become increasingly important for managing complex, long-running tasks and could evolve into a primary interface for computing.

OpenAI has been enhancing its voice features for years, with over 150 million people already using voice capabilities with ChatGPT. Rivals like Apple and Amazon are also developing more conversational AI assistants. OpenAI has incorporated safeguards to ensure age-appropriate responses and provide resources for sensitive topics, though some limitations, such as a noticeable American accent in a Hindi translation demo, were noted.

Frequently asked questions

GPT-Live refers to a new family of voice models from OpenAI that can listen and speak simultaneously in real time, enabling more natural conversations.

GPT-Live-1 mini will replace the current Advanced Voice Mode in ChatGPT for paid users, while the larger GPT-Live-1 model will also be available to them.

Full-duplex models allow for simultaneous speaking and listening, enabling natural interruptions and more fluid conversational exchanges.

What Happens Next

01GPT-Live-1 mini will be rolled out as the default for ChatGPT's Advanced Voice Mode.
02Users of paid tiers will gain access to the larger GPT-Live-1 model.
03OpenAI may explore hardware products like AI-enabled earbuds in the future.

Get the newsletter.

Pick the topics you actually care about. We'll email when there's news worth your time, on the cadence you choose. Cancel any time from your account.

Cadence

How It Developed

OpenAI launched GPT-Live, a new family of real-time voice models that can listen and speak simultaneously.
The new models, GPT-Live-1 and GPT-Live-1 mini, are designed for more natural conversations and better turn-taking.
GPT-Live-1 mini will replace the current Advanced Voice Mode in ChatGPT by default for paid users.
These full-duplex models enable users to interrupt naturally and support features like live translation.
The models can access newer GPT models for enhanced search, reasoning, and agentic capabilities.
OpenAI envisions voice as a primary interface for complex computing and long-running agentic work.
The live translation feature in Hindi demonstrated a heavy American accent and unnatural tone.

Sources

T1
OpenAI launches GPT-Live voice models that listen and speak simultaneouslyReuters
T1
OpenAI releases new voice models for more natural live conversationsTechCrunch
T2
OpenAI to publicly release GPT-5.6, rolls out Live voice AI modelscnbc.com

Related Stories

OpenAI's GPT-5.6 Model Approved for Broad Release by U.S. Government
8 Jul · 3:36 AM
BofA extends $520 million credit line to OpenAI ahead of IPO
8 Jul · 4:59 PM
OpenAI, SpaceXAI Announce New AI Models Amid Intensifying Rivalry
8 Jul · 11:00 AM
OpenAI, Meta, xAI Unveil New AI Models Amid National Security Review
8 Jul · 1:10 PM
Perplexity AI to use Nvidia's new CPUs for AI agent workloads
7 Jul · 7:05 PM