Supertone icon

Supertone

Interact with Supertone API for AI voice synthesis

Actions8

Speech → Predict Duration

AI-generated

Summary

Predicts speech duration for given text and voice without generating audio (free).

Inputs

  • voice (required) — Select the voice to use for speech generation (resource locator).
  • text (required) — The text to convert to speech (max 300 characters). Sent in request body as JSON property 'text'.
  • language — Language of the text. Defaults to Korean ('ko'). Sent in request body as JSON property 'language'. Options include en, ja, ar, bg, cs, da, nl, et, fi, fr, de, el, hi, hu, id, it, pl, pt, ro, ru, es, vi.
  • style — Speaking style (e.g., neutral, happy, sad, angry, kind-default, serene). If omitted, the voice's default style is used. Sent in request body as JSON property 'style' (undefined if empty).
  • requestOptions — Advanced request options (node settings). Includes batching configuration.
  • allowUnauthorizedCerts — Whether to accept responses with invalid SSL certificates.
  • proxy — HTTP proxy URL (supports basic auth).
  • timeout — Request timeout in milliseconds (minimum 1). Default 10000.

Output shape

object

Response is returned as JSON in the node output (json property). Typically includes a numeric field indicating predicted duration in seconds (e.g., { duration: 2.35 }). No binary data is produced.

Examples

Example 1: Estimate how long a Korean sentence will take when spoken with a specific voice.

Resource: Speech, Operation: Predict Duration, Voice: selected voice, Text: '안녕하세요, 오늘 날씨가 좋네요.', Language: ko, Style: neutral

Links

Discussion