GPT Live 1 and 1 mini power ChatGPT Voice globally
OpenAI rolled out its full-duplex GPT Live voice models (GPT Live 1 + GPT Live 1 mini) to all ChatGPT Voice users on iOS, Android, web, and CarPlay; simultaneous listening and speaking, reasoning delegation to backend models, live translation, visual cards.
Why it matters
Full-duplex conversation (0.8s turn-taking latency vs 1.4s for Realtime-2.1) with a two-model split: a cheap voice layer up front and frontier reasoning behind.
What you could build with it
A contact-center SaaS could rebuild its IVR on the GPT Live API via a provider like Twilio: natural interruption handling and low per-minute voice-layer pricing make high-fidelity phone agents economical, while complex tasks delegate to cheaper or stronger backends.
Does it hold up?
The API model has third-party benchmark and integration evidence (Twilio docs, Speak/Tau3 results); the ChatGPT Voice rollout itself is days old — too early for consumer-side quality verdicts.
Built with GPT Live 1 and 1 mini power ChatGPT Voice…
- OpenAI's GPT-Live-1 Listens and Speaks at the Same Timeaitechconnect.in · Independent analysis of the API model with published benchmarks (80.1% full-duplex interactivity, 87% tool-calling accuracy) and the voice-layer pricing model.
- Build Voice AI with Twilio GPT-Live-1 APItwilio.com · Production integration guide connecting GPT-Live-1 to Twilio Voice — evidence of real third-party adoption.
- OpenAI arms devs with AI conversation tool that can talk and listen at the same timetheregister.com · Early API-era reporting including the Speak evaluation (cut interruptions ~80% vs turn-based systems) and Tau3 benchmark score of 86.2%.
Learn more
First spotted on article: source.
More AI models
OpenAI publishes 722 AI-generated math manuscripts from an unreleased modelOpenAI released 722 mathematical manuscripts in 372 result families to GitHub under Apache-2.0, produced by an…models · JEV 0.74Mistral Large 4 (Le Chonk): 1.05T-param open-weight MoEMistral's new flagship: a 1.05T-parameter mixture-of-experts multimodal model with 49B active params and 1M-token…models · JEV 0.72ElevenLabs Eleven v4 + v4 Turbo: new emotive TTS architecture with a 100ms real-time variantElevenLabs released Eleven v4 and Eleven v4 Turbo on September 28: a new text-to-speech architecture with inline…models · JEV 0.72VoiceStudio: open-source, fully-local ElevenLabs alternative with voice cloning, dubbing and dictation in 646 languagesVoiceStudio is an open-source, fully-local voice AI studio: voice cloning from a 3-second sample, voice design, video…models · JEV 0.71JetBrains Mellum 2.1: 12B MoE coding model trained for agentsJetBrains released Mellum 2.1, a 12B MoE model (2.5B active parameters, 128K context) under Apache 2.0, trained mainly…models · JEV 0.69DeepSeek V4-Flash official API: public beta with upgraded agent capabilities, Responses API and Codex supportDeepSeek's official V4-Flash is now a public beta API with massively upgraded agent capabilities - benchmark scores…models · JEV 0.69
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.