# Live transcribe

The Live transcribe feature is a real-time speech recognition service that transcribes audio input on the fly. You can upload or stream audio documents for transcription through the [WebSocket API](https://speech-ntu.gitbook.io/speech-gateway/guide-1/live-transcribe/websocket-api).​

* [General Requirements](https://speech-ntu.gitbook.io/speech-gateway/guide-1/live-transcribe/general-requirements)
  * [Authentication](https://speech-ntu.gitbook.io/speech-gateway/guide-1/general-requirements#authentication)
  * [Language Model and Usage Limits](https://speech-ntu.gitbook.io/speech-gateway/guide-1/general-requirements#language-model-and-usage-limits)
  * [Data Privacy and Security](https://speech-ntu.gitbook.io/speech-gateway/guide-1/general-requirements#data-privacy-and-security)
  * [Supported Audio Formats](https://speech-ntu.gitbook.io/speech-gateway/guide-1/general-requirements#supported-audio-formats)
  * [Recording Guidelines](https://speech-ntu.gitbook.io/speech-gateway/guide-1/general-requirements#recording-guidelines)
* [WebSocket API](https://speech-ntu.gitbook.io/speech-gateway/guide-1/live-transcribe/websocket-api)
  * [Getting Started](https://speech-ntu.gitbook.io/speech-gateway/guide-1/websocket-api#getting-started)
  * [Transcribe a Live Recording](https://speech-ntu.gitbook.io/speech-gateway/guide-1/websocket-api#transcribe-a-live-recording)
  * [Transcribe an Audio File](https://speech-ntu.gitbook.io/speech-gateway/guide-1/websocket-api#transcribe-an-audio-file)
  * [End Transcription Session](https://speech-ntu.gitbook.io/speech-gateway/guide-1/websocket-api#end-transcription-session)
  * [Transcribe Audio and Save Adaptation State](https://speech-ntu.gitbook.io/speech-gateway/guide-1/websocket-api#transcribe-audio-and-save-adaptation-state)
  * [Server Response](https://speech-ntu.gitbook.io/speech-gateway/guide-1/websocket-api#server-response)
  * [Save Results](https://speech-ntu.gitbook.io/speech-gateway/guide-1/websocket-api#save-results)