Skip to content

LLM: Audio to Text

Field it applies to

  • Field type: string_long

File location

ai_automators/src/Plugin/AiAutomatorType/LlmAudioToStringLong.php

Description

LLM: Audio to Text allows AI-driven transcription of audio files into plain text. The transcribed text is stored in string_long fields, suitable for medium-length content.

Requirements

  • Field must be of type string_long.
  • The entity must have a field referencing audio files (audio/mpeg, audio/aac, audio/wav).

Configuration options

  • Base field selection: Choose the audio file reference field.
  • AI model selection: Select the AI model to use for transcription.

Behavior

  • Processes audio files from the configured base field.
  • Transcribes speech to text using an AI provider.
  • Stores the text in the target string_long field.
  • Validates that the output is a string.

Example use cases

  • Transcribe short audio memos into text.
  • Convert voicemail recordings into readable format.
  • Document short interviews or testimonials.
  • Generate text for short video captions.
  • Turn voice notes into structured text entries.
  • Create summaries from brief meeting audio clips.
  • Capture spoken instructions as written text.
  • Produce transcripts of social media audio clips.
  • Archive radio jingles or promos as text.
  • Enable search indexing of short audio snippets.

Notes

  • Raw AI output is stored directly in the field without exposing intermediate prompt data to users.
  • Only valid audio formats are accepted for processing.
  • Further validation or formatting logic can be added if required.

This documentation was AI-generated.