ModelOpenAIConfirmed
Story 2 of 16 in the Thursday 19 November editionOpenAI ships GPT-Live-1.1 voice API with $0.02/min standby
Live on 20 November 2026 over WebRTC and WebSockets with a 128,000-token context window, sharper speech transcription and $0.05 per active minute.

OpenAI released GPT-Live-1.1 (gpt-live-1.1) in the OpenAI API on 20 November 2026 for full-duplex voice agents that listen and speak at the same time over WebRTC or WebSocket transports. Compared with gpt-live-1, the model transcribes incoming caller speech more accurately, holds regional accents and voice identity more consistently across languages, and follows plain-language instructions on when to hand off reasoning or tool execution via client or Responses delegation alongside models such as GPT-6.2 Astra and Sol.
Context engine rollover, live timeline updates and per-second pricing
Each voice session defaults to a 128,000-token context window (accepting up to 16,384 tokens of startup instructions and up to 128 prior messages totalling 8,192 tokens) and automatically spins up a summarised replacement voice engine in the background when usage passes 90%. Applications inject up to 500 tokens mid-call through session.instructions.append, session.thinking.append and session.commentary.append, and set store: true to fork saved conversations for evaluation trials.
OpenAI bills gpt-live-1.1 per second at $0.05 per active minute and $0.02 per minute on WebRTC sessions while inference is paused in standby.
Technical details
- Model ID and transports
- gpt-live-1.1 over WebRTC (adaptive media track) and WebSocket
- Per-second session pricing
- $0.05 per active minute; $0.02 per minute in paused WebRTC standby
- Context window and startup limits
- 128,000 tokens total; up to 16,384 instruction tokens and 8,192 history tokens
- Mid-call append events
- session.instructions.append, session.thinking.append and session.commentary.append (500 tokens)
Release details & context
Full-duplex transcript deltas, microphone muting and voice options
During overlapping speech, GPT-Live-1.1 emits separate session.input_transcript.delta and session.output_transcript.delta events stamped with approximate start_ms and end_ms offsets from session start, allowing client interfaces to grow caller and assistant caption bubbles independently while running lightweight transcript guardrails on models like gpt-5.6-luna.
Sending session.input_audio.mute pauses caller capture while letting ongoing assistant speech and delegated backend tasks finish. Voice selection (marin by default, plus regional styles such as quartz and custom voices) and delegation mode are locked at session creation, with session.update scoped to Responses delegation settings.
Common questions
When does the $0.02-per-minute GPT-Live-1.1 standby rate apply?
The reduced $0.02-per-minute standby rate applies per second to WebRTC voice sessions during intervals when active inference is paused.
Sources
- PrimaryOpenAI API Docs: GPT-Live-1.1 Modeldevelopers.openai.com
- OfficialOpenAI API Docs: Managing GPT-Live sessionsdevelopers.openai.com
- OfficialOpenAI API Docs: Pricingdevelopers.openai.com
- On XOpenAI Developers (@OpenAIDevs) GPT-Live-1.1 launch post on Xx.com
OpenAI release timeline (56 launches tracked)
- OpenAI cuts Responses API background tokens up to 40% on GPT-6.2
OpenAI opens public-beta 2-hour and 8-hour Responses API background windows, cutting Standard token prices by 30% and 40% on GPT-6.2.
- OpenAI cuts AMD GPU PPA up to 14% and partners on GPT-Synopsys
OpenAI details up to 14% power, performance and area gains on AMD flagship GPU IP sent to manufacturing and announces GPT-Synopsys with Synopsys.
- OpenAI reports $2,400 daily median researcher agent spend in RSI study
OpenAI’s Jagged Frontier report logs $2,400 median daily researcher agent spend, 5.7 agent workdays per day and 77% Safety workweek growth.
- OpenAI rolls out real-time Live Dictation streaming inside ChatGPT
OpenAI ships Live Dictation in ChatGPT so spoken words stream into text in real time without post-recording transcription delays.

