OpenAI launches GPT-Live-1.1 voice model API
Understand this story
OpenAI has launched its full-duplex voice model GPT-Live-1.1 (model ID gpt-live-1.1) on the API, with a default context window of 128,000 tokens, accessible only through the OpenAI cloud via WebRTC or WebSocket, and cannot be deployed locally. Billing adds a new standby state: WebRTC idle time is billed at $0.02 per minute, lower than the $0.05 per minute when active, a 60% reduction; backend model and tool calls are billed separately. Switching models requires creating a new session.
AI overview based on source reports · Updated 11 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- MarkTechPostOpenAI Ships GPT-Live-1.1 for Voice Agents, Cuts Idle Cost to $0.02/Min
OpenAI released GPT-Live-1.1 in its API as gpt-live-1.1, a full-duplex voice model for multilingual conversations that can listen while speaking, with OpenAI claiming more natural accents, better transcription and more reliable task completion than GPT-Live-1, though no benchmark scores were published. The release adds a standby billing state that pauses inference during silence while the WebRTC connection stays open, billed at $0.02 per minute versus $0.05 active, with an idle timeout configurable from 5 to 60 seconds. The model is API-only with a 128,000-token default context window, supports audio and text input and output, and delegates reasoning and tool calls to a backend text model billed separately.
Activity for this story
- Current comparable activity
- 7
- Peak comparable activity
- 10Nov 21 21:00
- Change in the past 24 hours
- –
Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.