Introducing GPT-Live-1 in the API
OpenAI has released GPT-Live-1 via API, a full-duplex voice model priced at $0.05 per minute that handles listening and speaking within a single model, eliminating the latency and fragility of chained speech-to-text, LLM, and text-to-spe…
- 01The model outperforms GPT-Realtime-2.1 by 30 percentage points on Full Duplex Bench and ranks first on Tau3 for voice-agent intelligence when paired with GPT-6 Astra.
- 02Early enterprise deployments show measurable operational impact: Speak reduced tutor interruptions by approximately 80% versus prior turn-based systems.
- 03Developers can pair GPT-Live-1 with backend models of varying cost and reasoning depth—such as Luna for high-volume scheduling or Astra for complex reasoning tasks—and deploy across telephony, customer support, and business workflow use cases at scale.
OpenAI has released GPT-Live-1 via API, a full-duplex voice model priced at $0.05 per minute that handles listening and speaking within a single model, eliminating the latency and fragility of chained speech-to-text, LLM, and text-to-speech architectures. The model outperforms GPT-Realtime-2.1 by 30 percentage points on Full Duplex Bench and ranks first on Tau3 for voice-agent intelligence when paired with GPT-6 Astra. Early enterprise deployments show measurable operational impact: Speak reduced tutor interruptions by approximately 80% versus prior turn-based systems.
Read the full article at openai.comOpenAI has released GPT-Live-1 via API, a full-duplex voice model priced at $0.05 per minute that handles listening and speaking within a single model, eliminating the latency and fragility of chained speech-to-text, LLM, and text-to-speech architectures. The model outperforms GPT-Realtime-2.1 by 30 percentage points on Full Duplex Bench and ranks first on Tau3 for voice-agent intelligence when paired with GPT-6 Astra. Early enterprise deployments show measurable operational impact: Speak reduced tutor interruptions by approximately 80% versus prior turn-based systems. Developers can pair GPT-Live-1 with backend models of varying cost and reasoning depth—such as Luna for high-volume scheduling or Astra for complex reasoning tasks—and deploy across telephony, customer support, and business workflow use cases at scale.
Don't miss tomorrow's
The Daily Pulse in your inbox each morning — sourced and linked.