4Jawaly Actions

4Jawaly utility actions and helpers

Actions9

Transcribe Audio

AI-generated

Overview

This node provides multiple utility actions including the ability to transcribe audio files into text using the OpenAI Whisper API. It supports various audio sources such as file paths, URLs, and binary data from previous nodes. The node is particularly useful for converting speech to text with excellent support for Arabic language, and it offers options for language detection, response format, and automatic translation of the transcribed text into Arabic and English.

Use Case Examples

  1. Transcribing a local audio file to text for creating subtitles or transcripts.
  2. Converting an audio file from a URL into text for content analysis or archiving.
  3. Using binary audio data from a previous node to transcribe speech in a workflow.

Properties

Name Meaning
Service The transcription service to use, currently supports OpenAI Whisper API with excellent Arabic support.
Audio Source The source of the audio file to transcribe, can be a file path, URL, or binary data from a previous node.
File Path Full path to the audio file on the device, required if Audio Source is 'File Path'.
Audio URL Direct URL to the audio file, required if Audio Source is 'URL'.
Binary Property Name of the binary property containing the audio data, required if Audio Source is 'Binary Data'.
Language Language of the audio file to improve transcription accuracy; can be auto-detected, Arabic, or English.
Response Format Format of the transcription result, can be plain text, JSON with additional details, or verbose JSON with timestamps.
Auto Translate If enabled, automatically translates the transcribed text into Arabic and English.

Output

JSON

  • success - Indicates if the transcription was successful.
  • fileName - Name of the audio file processed.
  • language - Language code used for transcription.
  • text - Transcribed text output (if response format is text).
  • transcriptions - Detailed transcription data in JSON or verbose JSON formats.
  • translations
    • original_language - Detected original language of the transcribed text when auto-translate is enabled.
    • original_text - Original transcribed text before translation.
    • arabic_text - Translated text in Arabic.
    • english_text - Translated text in English.

Dependencies

  • OpenAI API key credential for authentication to use OpenAI Whisper and GPT-3.5-turbo models.

Troubleshooting

  • Ensure the audio file path or URL is correct and accessible; file not found errors occur if the path is invalid.
  • Binary data must be correctly provided in the specified binary property; missing or incorrect binary data will cause errors.
  • OpenAI API key must be configured properly; authentication errors will occur without valid credentials.
  • The node throws errors if required parameters like file path, URL, or binary property are missing depending on the audio source.
  • If auto-translate is enabled, translation failures may occur due to API limits or network issues; these are reported in the output.

Links

Discussion