Skip to main content
工作流应用

Audio Parsing Node

The Audio Parsing Node converts audio files into structured text, outputting speech-to-text (ASR) transcriptions along with their corresponding time-stamped information. The parsed results can be further processed and analyzed by downstream nodes—such as Large Language Model (LLM) Nodes. This feature is currently free of charge.

Usage Example

For example, to parse an audio file and generate a content summary:
This workflow accepts an audio file, extracts ASR text and timing information via the Audio Parsing Node, and then passes the result to an LLM Node to generate a concise audio summary.

Adding the Node

Method 1: Add via the + Button

  • Hover over any node on the canvas (e.g., the Start Node); a + button automatically appears on its right side.
  • Click the + button, select Data > Audio Parsing from the pop-up node list. The node is added to the canvas and automatically connected to the upstream node.

Method 2: Drag from the Node Library

  • In the left-hand node library, locate and select Data > Audio Parsing.
  • Drag the node onto the canvas at your desired location, then drag a connection line from the edge of an upstream node to the Audio Parsing Node.

Core Configuration

Input

Specify the audio file to be parsed. You may either directly input a URL or reference a variable from an upstream node. Use the dropdown to choose the input method:
MethodDescriptionUse Case
ReferenceReference a variable from an upstream nodeAudio is provided as a String-type URL or a File-type object from the Start Node or another upstream node.
InputDirectly enter an audio URLAudio URL is static or obtained externally.
Limitations:
  • Maximum file size: 512 MB
  • Only single-file parsing is supported (no batch or list-based parsing)
  • Supported formats: .aac, .amr, .flac, .flv, .m4a, .mp3, .mpeg, .ogg, .opus, .wav, .webm, .wma

Parser Selection

Parser TypeBillingDescription
LLM-based Audio/Video ParserFree of charge (for now)Uses built-in large language models for parsing; prompt customization is not supported.

Output Variables

The Audio Parsing Node outputs structured ASR results based on the audio content, including file type and time-segmented transcription.
Variable NameTypeDescription
fileTypeStringFile extension (e.g., "wav", "mp3").
segmentsArray<Object>Array of time-segmented transcription results; each element represents one audio segment.
Structure of each segments array element:
Field NameTypeDescription
audioFramesArray<Object>Array of audio frames containing ASR details for this segment.
startTimeNumberStart time of the segment (in milliseconds).
endTimeNumberEnd time of the segment (in milliseconds).
indexNumberSegment index (0-based).
Structure of each audioFrames array element:
Field NameTypeDescription
ASRInfoStringSpeech-to-text transcription (ASR result).
startTimeNumberStart time of the audio frame (in milliseconds).
endTimeNumberEnd time of the audio frame (in milliseconds).

Common Usage Patterns

Paired NodeTypical Use Case
Start Node → Audio Parsing NodeThe Start Node provides an audio file (either a File-type object or a String-type URL), and the Audio Parsing Node transcribes its content.
Audio Parsing Node → LLM NodeThe Audio Parsing Node outputs structured data (e.g., ASR text from segments), which the LLM Node uses to generate summaries, answer questions, or extract key information.
Audio Parsing Node → Variable Processing NodeExtract specific fields from the parsing result for formatting or data transformation (e.g., concatenate all ASRInfo values into a single transcript).