Azure AI icon

Azure AI

Call Azure AI Cognitive Services — Document Intelligence, Content Safety, Translator, and Language

Actions21

Azure OpenAI → Transcribe Audio

AI-generated

Summary

Transcribe an audio file to text using the Azure OpenAI Whisper deployment, allowing optional specification of language and response format.

Inputs

  • Authentication (required) — Choose between API Key or Microsoft Entra ID (OAuth2) authentication methods.
  • Deployment (required) — Select the Azure OpenAI deployment for audio transcription, with presets for Whisper v2, Whisper v1, or specify a custom deployment name.
  • Custom Deployment Name — Provide the exact name of your custom Azure OpenAI deployment if 'Custom' is selected.
  • Binary Property (required) — Name of the binary property from a previous node containing the audio file to transcribe.
  • Language (Optional) — ISO 639-1 language code for the audio. Leave blank to enable automatic language detection.
  • Response Format (required) — Choose the transcription output format: text-only JSON, verbose JSON with timestamps, plain text, SRT subtitles, or VTT subtitles.

Output shape

A single JSON object representing the transcription result or text string depending on format.

The output JSON includes the transcription text and optionally detailed metadata such as timestamps if verbose JSON is selected. If a non-JSON format is chosen, raw text or subtitle content is returned under the 'text' property.

Examples

Example 1: Transcribe an audio file stored in a binary property called 'audioData' with auto language detection and receive a simple JSON text output.

Set 'audioBinaryProperty' to 'audioData', leave 'audioLanguage' empty, choose 'whisper' deployment and 'json' response format.

Example 2: Transcribe audio with specified English language code and verbose JSON output including timestamps for word alignments.

Set 'audioLanguage' to 'en', select 'whisper' deployment, and 'verbose_json' response format.

Links

Discussion