Event ·Multi-source

OpenAI 推出 GPT-Live-1 实时语音模型 API，支持打断处理与工具调用

OpenAI 发布 GPT-Live-1 实时语音模型 API，该模型支持全双工对话，能同时听取和输出语音，并处理打断、停顿及背景噪声。它将语音理解与输出整合，减少传统语音转文字和文字转语音的延迟。在 Full Duplex Bench 测试中，GPT-Live-1 表现优于 GPT-Realtime-2.1，在 Tau3 任务测试中排名第一。开发者可通过系统提示词调整其语气和语速。

Current read

OpenAI 推出了 GPT-Live-1 实时语音模型 API，支持打断处理和工具调用，能帮你快速构建全双工语音应用，比之前的方案更高效。

11 sources78° AI heatLast updated 2026/9/11 00:28:22

Evidence

11 sources

Primary sourceIT之家

2026/9/11 00:28:22View source

Related source 1@OpenAIDevs

2026/9/10 18:35:47View source

Related source 2Decoder

2026/9/10 17:47:35View source

Related source 3HeyGen

2026/9/10 17:55:10View source

Related source 4OpenAI Blog

2026/9/10 00:00:00View source

Related source 5Genspark

2026/9/10 18:11:32View source

Related source 6Thomas Wolf

2026/9/10 18:57:00View source

Related source 7Greg Brockman

2026/9/10 21:13:24View source

Related source 8shao__meng

2026/9/9 00:30:22View source

Related source 9爱范儿

2026/9/9 02:14:52View source

Related source 10歸藏(guizang.ai)

2026/9/9 02:32:45View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story