# Transcribe audio to text (Whisper)

Whisper is a speech-to-text recognition model

* You can upload your audio in a few ways:
  * From the Internet: Provide a direct link (URL) to the audio file.
  * From Your Computer: Upload the audio file directly from your device.
* File String: Upload the audio in string format if needed.
* Priority Queue: Set the order in which your request is processed.
* File URL: Enter the direct URL of the audio file.
* Group Segments: Decide if you want to group transcription segments together.
* Transcript Output Format:
  * Words: Get a plain text transcript containing only words.
  * Segments: Get a transcript with timestamps for creating subtitles.
* Num Speakers: Specify the number of speakers, or let Whisper automatically detect them.
* Language: Select the language of the audio file, or let Whisper automatically detect it for you.
* Prompt: Add any specific words or phrases used in your audio to make the transcription more accurate.

Offset Seconds: Set a starting point if you need the transcription to begin at a specific point in the audio.