GPT-Live-1 does full-duplex listening and speaking in one model; open-source demo with LiveAvatar
HeyGen · x · 2026-09-11
TryLiveAvatar, a launch partner for OpenAI's GPT-Live-1, built and open-sourced a full demo. GPT-Live-1 is a single model that listens and speaks simultaneously — no STT→LLM→TTS chain and no brittle handoffs. It handles interruptions, background noise and mid-sentence changes of mind, offloading deeper reasoning to a backend model while the conversation keeps flowing. Paired with LiveAvatar's real-time avatars, full-duplex audio plus a live face is claimed to be the closest AI gets to talking to a person.
More from Multimodal
- Wan2GP lands on Pinokio: 1-click AI video on 6GB GPUs — cocktailpeanut · 2026-10-27
- WanGP v18.81 adds Tencent's Hy-Nomas world model, playable on 8GB VRAM — cocktailpeanut · 2026-10-27
- World models matter even without realtime: closed loops and persistent memory — cocktailpeanut · 2026-10-27
- Butter.video wins over skeptical creator with hundreds of tasteful timeline 'blocks' for video effects — beechinour · 2026-10-27
- AI Video Pits Brock Lesnar Against Duncan the Tall in a Toy Showdown — KloneStudios · 2026-10-27
- $50, 2-hour Kurzgesagt-style AI video reignites debate over intent in AI creation — motionbynick · 2026-10-27