GPT-Live
GPT-Live is OpenAI's voice model featuring full-duplex architecture that enables simultaneous listening and speaking for real-time bidirectional audio interactions with latency under 500 milliseconds. The model is available on iOS, Android, and ChatGPT.com, supporting use cases including live tutoring, language practice, and interview preparation, while delegating complex reasoning tasks to GPT-5.5 to maintain natural conversation flow.
Updated 3 August 2026
Specifications
No specifications recorded yet.
Latest developments
The Sequence Radar #893: Last Week in AI: GPT-5.6, Grok 4.5, Muse Spark 1.1 and the Post-Chatbot Stack
TheSequence · 3 weeks ago ·
39
Deep Learning Weekly: Issue 463
Deep Learning Weekly · 3 weeks ago ·
8
Introducing GPT-Live
TLDR Dev · 3 weeks ago ·
39
GPT-Live Enables New Use Cases Like Live Tutoring and Real-Time Language Practice
The Neuron · 3 weeks ago ·
22
GPT-Live Now Available on iOS, Android, and ChatGPT.com
The Neuron · 3 weeks ago ·
2
Introducing GPT‑Live
Simon Willison · 3 weeks ago ·
50
Introducing GPT-Live
OpenAI Blog · 3 weeks ago ·
35
2026
OpenAI Developing Screenless AI Speaker with Movement Capabilities as First Hardware Device Product launch
OpenAI releases GPT-Live, a low-latency voice model for real-time conversational AI Feature update
OpenAI releases GPT-5.6 in multiple variants alongside frontier AI model releases from competitors Model release
- OpenAI's first device is reportedly a screenless ChatGPT speaker with cameras and movement
- The Sequence Radar #893: Last Week in AI: GPT-5.6, Grok 4.5, Muse Spark 1.1 and the Post-Chatbot Stack
- Deep Learning Weekly: Issue 463
- Introducing GPT-Live
- GPT-Live Enables New Use Cases Like Live Tutoring and Real-Time Language Practice
- GPT-Live Now Available on iOS, Android, and ChatGPT.com
- Introducing GPT‑Live
- Introducing GPT-Live