OpenAI Rebuilds ChatGPT Voice With GPT-Live to Talk and Listen at Once
OpenAI's GPT-Live gets a full engineering teardown: WARP protocol, stateful inference, and async delegation make voice feel truly live
- GPT-Live goes full-duplex: OpenAI's new voice system listens and speaks simultaneously, removing the turn detector that caused unnatural interruptions.
- Two-model architecture: A lightweight voice model handles conversation flow while GPT-5.5 handles search and reasoning asynchronously in the background.
- WARP protocol: OpenAI built and open-sourced a new WebRTC extension that cuts session startup from 6 network round trips to 1, now submitted to the IETF.
- Stateful inference with seamless handoffs: Context compaction and model instance transitions happen without pausing the audio stream.
- Availability: GPT-Live-1 mini is free; full GPT-Live-1 requires Go ($8/mo) or above. API access is not yet available.
- No video or screen sharing at launch: Legacy Advanced Voice Mode remains available for those features while GPT-Live catches up.
OpenAI published a detailed engineering post on how they rebuilt ChatGPT Voice to power GPT-Live, their third-generation voice system. The headline feature is full-duplex audio — the model can listen and speak simultaneously — but the deeper story is a complete rearchitecture of every layer between the user's microphone and the model, built over six months to serve 150 million weekly voice users.
Why the old approach broke down
Every previous ChatGPT voice system was turn-based. Advanced Voice Mode processed and generated audio within a single model, which reduced latency and smoothed conversations, but the model still had to wait for the user to stop speaking before it could respond. A small "turn detector" model decided when the user had finished talking. Because that detection relied on silence, a brief pause or background noise could trigger a premature response, cutting the user off at unnatural moments.
Two architectural bets
GPT-Live is built on two structural changes that work together:
- Full-duplex continuous inference: Instead of processing a sequence of separate messages, GPT-Live continuously processes input while generating output. The model makes interaction decisions many times per second: whether to speak, keep listening, pause, interrupt, or invoke a tool.
- Asynchronous delegation: The voice interaction layer is decoupled from the reasoning layer. Straightforward questions are handled directly by GPT-Live. Queries that require web search, deeper reasoning, or agentic work are delegated to a frontier model running in the background — GPT-5.5 at launch — while GPT-Live keeps the conversation going.
That continuous-processing design enables live translation and lets the model acknowledge a speaker mid-sentence with cues like "mhmm," rather than waiting for a full turn to end.
Rebuilding the media pipeline
Keeping audio flowing without gaps required rethinking every layer of the stack. OpenAI made a hard separation between the media path and everything else, then rewrote the media frontend and inference logic in Go, replacing a previous Python