Version 4

OpenAI launched GPT-Live on July 8, 2026, replacing Advanced Voice Mode with a full-duplex architecture that listens and speaks simultaneously while delegating complex tasks in the background to GPT-5.5 [1]. GPT-Live-1 serves paid tiers (Go, Plus, Pro); GPT-Live-1 mini serves free users [3]. The only extended independent review — from Simon Willison — is positive but documents a behavioral bug that shipped and required a user report to fix [4]. Coverage continues to accumulate but adds no new substantive analysis.

Full-duplex voice removes the turn-taking constraint that made prior voice AI feel mechanical. If background delegation to GPT-5.5 holds at scale, it offers voice-native access to frontier reasoning without sacrificing conversational responsiveness. The ambient interruption concern points to a real deployment question about where full-duplex is appropriate — focused one-on-one sessions versus open or shared environments.

On July 8, 2026, OpenAI launched GPT-Live, a voice AI built on full-duplex architecture that allows the model to listen and speak simultaneously, replacing the previous GPT-4o-era Advanced Voice Mode [1]. The prior system used discrete turn-taking — the model waited for the user to finish before responding — which many found stilted [2]. GPT-Live eliminates that constraint by processing incoming speech while generating output, handling interruptions in real time, and producing active-listening cues during conversation. GPT-Live-1 is the default for paid tiers (Go, Plus, Pro); GPT-Live-1 mini serves free users [3].

The system uses a two-tier design: a lightweight voice model maintains the conversational surface while tasks requiring web search, deeper reasoning, or agentic work are delegated in the background to GPT-5.5, with results returned without disrupting flow [1][4]. OpenAI benchmarks GPT-Live-1 above its predecessor on GPQA (expert-level scientific reasoning) and BrowseComp (agentic web search) [1], and cites more than 150 million weekly ChatGPT voice users as deployment context. Voice-specific real-time safety mechanisms can steer, modify, or end a conversation when potentially unsafe output is detected, with dedicated protections for teen users [1].

The only extended independent review comes from Simon Willison, who had preview access for several weeks. He describes GPT-Live as a credible brainstorming partner during hour-long sessions and confirms the delegation mechanism functions as described [4]. He also documented a behavioral bug — the model repeatedly interrupted with laughter at non-humorous statements — that OpenAI addressed after he filed a report [4]. His overall assessment is positive; he had largely stopped using Advanced Voice Mode due to its weaknesses, making his return to the feature a concrete if measured endorsement.

One concern to emerge from initial coverage is whether full-duplex interruption suits all environments. A newsletter review flagged that the model might interject too readily in ambient listening or meeting contexts — a usability tradeoff the launch framing did not address [2]. The same coverage also reports that OpenAI plans to release GPT-5.6 soon as its last 5.x release, with GPT-6 expected roughly a month later on a significantly larger pretraining base [2], providing context for where GPT-Live sits in OpenAI's near-term model progression.