OpenAI launches GPT‑Live‑1.1 voice model with cheaper standby billing
GPT‑Live‑1.1 adds full‑duplex multilingual speech, standby mode at $0.02/min, and runs only on OpenAI’s cloud.
The model supports full‑duplex audio, allowing it to listen and speak in a single stream.
Live sessions use a default 128,000‑token context window and connect via WebRTC or WebSocket. No open weights or quantized versions are available.
The release introduces a standby billing state that pauses inference during silence. Standby costs $0.02 per minute, compared with $0.05 per minute for active sessions.
GPT‑Live‑1.1 improves accent naturalness, transcription accuracy, handling of letters and numbers, and task completion reliability, according to OpenAI’s self‑reported claims.
The model remains API‑only, runs on OpenAI’s cloud, and does not support image, video, or structured outputs. Backend tool usage is billed separately.
Ce que cela signifie pour vous
The lower standby rate reduces cost for voice agents that spend time waiting for user input, making continuous‑listening applications more affordable. Because the model is cloud‑only and closed‑source, users seeking free or locally hosted alternatives must look to other open models rather than GPT‑Live‑1.1.
Sources
- OpenAI Ships GPT-Live-1.1 for Voice Agents, Cuts Idle Cost to $0.02/Min, MarkTechPost, 21 Nov 2026
- Managing GPT-Live sessions | OpenAI API, developers.openai.com (official), 21 Nov 2026
Rédigé automatiquement à partir des sources ci‑dessus, puis vérifié par notre code : chaque nombre et chaque nom doivent y figurer, et aucune phrase ne doit copier une source. Une erreur ? Signalez‑la.
Quelle IA pouvez‑vous utiliser gratuitement aujourd’hui ?
Tous les modèles IA gratuits, classés par votes humains anonymes, avec la mémoire et la vitesse dont ils ont besoin sur votre propre GPU.