Audio Transcription
AI-generatedSummary
Transcribe and optionally summarize audio files provided as binary data within the workflow.
Inputs
- System Instruction — Optional instructions that define the assistant behavior or personality during transcription.
- Temperature — Number between 0 and 2 controlling the randomness of the transcription output.
- Top P — Number between 0 and 1 controlling nucleus sampling for output token selection.
- Max Output Tokens — Maximum number of tokens in the transcription model response.
- Instruction — Text instruction guiding how the audio should be processed (e.g., transcribe and summarize).
- Binary Property — Name of the binary property in the input item that contains the audio file; defaults to 'data'.
Output shape
A single JSON object per input item containing the transcription and summarization result.
Each output corresponds to an input item; if an input item lacks the specified binary audio property, an error is thrown unless 'Continue On Fail' is enabled, in which case the error message is returned instead.