MiniMax TTS icon

MiniMax TTS

Convert text to speech with MiniMax TTS API

MiniMax TTS

AI-generated

Summary

Synthesize speech from text via MiniMax TTS API, returning optional binary audio and API response metadata.

Inputs

  • model — TTS model used to synthesize speech
  • text — Text to convert to speech. If empty, it will be read from upstream JSON (text, output, response, or content).
  • stream — Whether to enable streaming response
  • voiceId — Voice preset ID for synthesis
  • voiceSettings — Voice setting options: speed (default 1), vol (default 1), pitch (default 0), emotion (default happy)
  • audioSettings — Audio setting options: sampleRate (default 32000), bitrate (default 128000), format (default mp3), channel (default 1)
  • pronunciationTone — Optional tone dictionary entries, one per line, for example: process/(chu3)(li3)
  • subtitleEnable — Whether to include subtitle data in response
  • outputFormat — Audio output format for non-streaming mode
  • returnBinary — Whether to return audio as binary data on property audio

Output shape

json: API response object with audio and audio_url fields removed and trace_id attached; binary: binary audio data when returnBinary is true and stream is false, containing audio buffer with filename and MIME type

When stream is false and returnBinary is true, the node downloads the audio URL if present, or decodes a hex string, and attaches the resulting buffer as binary property audio. The JSON output always includes trace_id and excludes audio fields.

Examples

Example 1: Basic text-to-speech with default model and binary audio output

Set text parameter; leave model, voiceId, and audioSettings at defaults; ensure stream is false (default) and returnBinary is true (default).

Links

Discussion