Actions25
- Post Actions
- Topic Actions
- Upload Actions
Post → Generate TTS Audio
AI-generatedSummary
Generates speech audio from text using ElevenLabs, optionally creating transcript alignment data and uploading the result to DigitalOcean Spaces.
Inputs
- cleanedText (required) — The text to synthesize (e.g., from Prepare for TTS output).
- voiceId (required) — The ElevenLabs voice ID to use.
- modelId (required) — The ElevenLabs model ID (defaults to eleven_multilingual_v2).
- audioFormat — The final output format (MP3 or WAV). WAV is recommended for deterministic stitching.
- chunkStrategy — How to split the input text into chunks for synthesis (Paragraph Aware or Sentence Aware).
- resolveModelLimitFromApi — Whether to dynamically query ElevenLabs for the model's character limit.
- includeBinaryData — Whether to return generated audio/transcript files as n8n binary output.
- audioBinaryFieldName — The output field name for the audio binary data.
Output shape
A single item containing binary audio data (if enabled) and JSON metadata (transcript, alignment, diagnostics).
If the text exceeds the model's character limit, the operation splits the text into chunks, synthesizes them in parallel, and stitches the audio. Includes chunk-level diagnostics in JSON output. If configured, uploads the final audio to DigitalOcean Spaces.
Examples
Example 1: Convert a cleaned blog post into an MP3 audio file using a specific voice.
Use 'Paragraph Aware' chunking, MP3 format, and include binary output.
Example 2: Generate an audio file with precise timestamps for video editing.
Use WAV format (recommended for sync), 'Sentence Aware' chunking, and enable transcript alignment JSON.