Recently, OpenAI introduced a new feature called "ChatGPT Record" in its macOS desktop application. This feature is designed specifically for Pro, Team, Enterprise, and Edu users, enabling them to record audio, transcribe it, and generate audio summaries. Users can record up to 120 minutes of audio per session, and the recordings will automatically generate structured summaries.

The "ChatGPT Record" function aims to enhance user productivity, especially when handling meeting records, interviews, or lectures. Users only need to start the recording, and the system will simultaneously transcribe the audio in real-time while generating a concise summary afterward. This not only saves time from manual note-taking but also helps users better organize and understand the information.

According to OpenAI, this tool performs best with English recordings, ensuring accuracy in transcription and summarization. Additionally, all audio recordings are automatically deleted after transcription to protect user privacy. OpenAI further emphasizes that these audio data are not used for model training, providing additional security and trust for users.

Enterprise and education users can access this feature via the Compliance API to ensure compliance during usage. Since its launch in early June, ChatGPT Record has garnered significant attention and positive feedback from users, many of whom have expressed that this feature has made their work more efficient and convenient.

OpenAI's new feature not only demonstrates its technical prowess in audio processing but also provides users with a novel solution for information acquisition and management.

Key points:

📅 Each recording can last up to 120 minutes, and a structured summary is automatically generated after recording.  

🔒 All recordings are automatically deleted after transcription, ensuring user privacy and security.  

🌐 The feature currently works best for English speakers and is accessible to enterprise and education users via the Compliance API.