Event ·Official first-hand

OpenAI API 新增 GPT-Live-Transcribe 和 GPT-Transcribe 两个转录模型

OpenAI 在 API 中推出两个新转录模型：GPT-Live-Transcribe 专为低延迟实时转录设计，GPT-Transcribe 优化了已完成音频文件的异步转录和批量处理。这两个模型能更好理解上下文，在口音、多种语言、短句、数字、专业术语以及嘈杂背景下的语音识别准确率更高。

Current read

OpenAI 出了两个新转录模型，一个实时低延迟，一个异步批量处理，对口音和噪音场景更准。做语音识别应用的话可以试试。

11 sources51° AI heatLast updated 2026/7/28 18:01:09

Evidence

11 sources

Primary source@OpenAIDevs

2026/7/28 18:01:09View source

Related source 1IT之家

2026/7/28 23:01:33View source

Related source 2小互

2026/7/29 09:25:04View source

Related source 3Decoder

2026/7/29 12:45:33View source

Related source 4OpenAI

2026/7/27 17:32:29View source

Related source 5Greg Brockman

2026/7/27 23:00:02View source

Related source 6歸藏(guizang.ai)

2026/7/28 02:53:49View source

Related source 7AI Will

2026/7/28 07:00:20View source

Related source 8Latent Space (swyx)

2026/7/28 15:26:30View source

Related source 9LangChain

2026/7/28 19:59:57View source

Related source 10Latent.Space

2026/7/28 21:30:00View 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