OpenAI launches GPT-Live voice model that listens and speaks simultaneously
Technology
OpenAI just launched GPT-Live, its most advanced voice model so far.
It can listen and talk at the same time, making conversations feel way more natural: think less awkward pausing and more real-life flow, with little cues like "mhmm" and "yeah" thrown in.
GPT-Live supports interruptions and web searches
You can interrupt, pause, or ask it to slow down on the fly. It handles web searches and complex tasks behind the scenes, then brings those answers right into your chat.
GPT-Live keeps up with you without missing a beat, even through background noise, and can show info cards for stuff like weather or maps.
It's available in two versions (regular and Mini), with API access coming soon. User tests say it's faster, more attentive, and just plain easier to use than older models.