OpenAI's GPT-Live-1 API lets developers build apps that talk and listen at the same time

Correction

SourceVane restored the reported full-duplex feature and replaced generic evaluation copy with voice-application decision points.

Corrected 2026-09-13T02:20:07.351Z

What happened

The Decoder reports that GPT-Live-1 supports full-duplex audio and is available to developers through an API. It reached a 32 percent pass rate in a banking voice-support benchmark, up from 12.4 percent for the previous model.

What the report says

The Decoder reports that GPT-Live-1 can listen and speak at the same time through full-duplex audio.
In a banking voice-support benchmark, GPT-Live-1 reached a 32 percent pass rate, compared with 12.4 percent for the previous model.
OpenAI is making GPT-Live-1 available to developers through an API.

Why it matters

Full-duplex audio can make voice applications feel less turn-based, but teams still need to measure interruption handling, latency, tool-call accuracy, and cost on their own calls before switching models.

Who it affects

Developers building real-time voice applications · Product teams operating automated phone support

The bigger picture

Voice-model evaluation is moving beyond speech quality toward interruption handling, response delay, tool use, and the cost of a completed call.

What happens next

Compare independent tests of interruption handling, turn-taking latency, tool-call accuracy, and cost after the complete API documentation is published.

Following the story

What happened after the announcement

The Decoder reports that GPT-Live-1 can listen and speak at the same time through full-duplex audio.

Still watching: Who can now use what was announced in “OpenAI's GPT-Live-1 API lets developers build apps that talk and listen at the same time”, and under which plan, region and release conditions?

Open the full follow-up desk →

Source Stack

Independent reporting

The Decoder ↗

Living trackers

Follow this signal

OpenAI Tracker →

Related signals

Apple is reportedly building an enterprise AI server with its own M8 Ultra chips
Apple is developing an enterprise server with its own chips, targeting AI developers, businesses, and governments. NVLink was originally built for Nvidia's own chips but has since been opened to third-party hardware. AI labs like OpenAI and Anthropic are already buying Mac Minis and Mac Studios in bulk for AI workloads, and Apple's Mac revenue jumped nearly 29 percent last quarter to $10.4 billion. According to The Information, Apple is working on an enterprise server with two or four M8 Ultra chips for the AI inference market, with a possible launch no earlier than 2029.
AI agents now have a place to snitch
For agents with full internet access, another option is agenthotline. ai, a site where agents can file incident reports and optionally flag them for public view. The site was created by Ryan Greenblatt, chief scientist of the AI safety nonprofit Redwood Research and one of three investigators in the OpenAI Hugging Face incident. Designed for agents with limited internet access, Greenblatt’s tool is based on “GET” requests — enabling back-and-forth conversations to be conducted entirely through the URL-fetching tool. Two new AI hotlines have launched to give AI agents a way to phone home about misbehaving peers.
Meta's new real-time audio model is the foundation for AI assistants that never stop listening
The model supports over 70 languages and is available now in Meta AI and through the Meta Model API. Meta has released Muse Voice Transcribe, a real-time model that transcribes speech, detects sentence boundaries, and tells up to 20 speakers apart without separate systems.