Speech to Text (Whisper)

Audience: Citizen Developers

Skill Prerequisites: Tokens

This action extracts text from an audio file using the OpenAI Whisper API.

Typical Use Cases

• Convert speech to text in an audio file
• Transcribe audio recordings into written content

Don't use it to

• Extract text from non-audio files
• Transcribe audio content not supported by OpenAI Whisper

Related Actions

None

Input Parameter Reference

Parameter	Description	Supports Tokens	Default	Required
OpenAi API Key	The API key for Open AI.	Yes	empty string	Yes
Audio file	Identifier for the audio file. Can be: RelativeUrl, AbsoluteUrl, LinkClickUrl, or PhysicalPath.	Yes	empty string	Yes
Response Format	Choose from the following: json, text, srt, verbose_json, vtt.	Yes	json	No
Prompt	Optionally, provide additional instructions for OpenAi to use while creating the transcript.	Yes	empty string	No

Output Parameters Reference

Parameter	Description
Output Token Name	Stores the extracted text in a token with the specified name.

Examples

1. Basic Example

Extract text from an audio file and store the result in a token named "Transcript".

[
  {
    "Id": -1,
    "ActionErrorMessage": null,
    "ActionType": "OpenAi.Whisper",
    "Condition": null,
    "Description": null,
    "IsDeleted": false,
    "EventName": "",
    "Parameters": {
      "ApiKey": "your_openai_api_key",
      "AudioFileIdentifier": "path/to/your/audio/file.mp3",
      "ResponseFormat": {
        "IsExpression": false,
        "Expression": null,
        "Value": "text"
      },
      "Prompt": null,
      "OutputTokenName": "Transcript"
    }
  }
]

For example, you can import this JSON content into a workflow using the "Import Actions" button.